ProgrammableMarketing/avallon.ai (opens in a new tab)
PDF

Views

  • Plan
  • Roadmap
  • Research

On this page

  • The brief
  • Start here
  • Idea portfolio
  • Effort vs impact
  • This week · 6
  • This quarter · 11
  • Big bets · 3
  • Wildcards · 2
  • How we can help
Generated Nov 10, 2026
  1. Programmable
  2. Playbooks
  3. Avallon
Avallon logo

growth playbook · Nov 10, 2026

avallon.ai (opens in a new tab)

Avallon growth playbook

Diagnosis

Avallon is a 2025 Y Combinator company with a $4.6 million seed round. It went from zero to seven figures of ARR in twelve months and is growing 3X quarter over quarter. It has strong, specific proof from TPAs: 64% of calls resolved at Athens Administrators, 90% of intake at Johnston & Associates and 54,000 medical reports a year at JAG Source.

Its job posts say the next step is carriers: the largest insurers will run claims through the product within months, and the team plans to more than double. The website has not caught up. It has 24 pages, no blog, pricing, integration detail or downloadable proof. All proof is workers' comp, yet the life, P&C, specialty and carrier pages show mockups. Every demo goes to one person's calendar.

The carrier AI review that the site promises to pass is never mapped to the NAIC AI Model Bulletin, which insurers must now apply to their vendors. This playbook first arms the carrier buying committee: a regulatory vendor pack, sourced numbers, line-specific proof, shareable case studies and a routed demo request. Next come integration pages, partner programmes, published pricing logic, a packaged 40-claim proof of value and TPA-focused outbound. The big bets are a claimant portal carriers can buy first, an annual benchmark from agent data and a European TPA beachhead.

Growth Score

7/10

Biggest Opportunity

Carriers are the next market, and their buying committees decide on regulatory, security and integration evidence. Avallon already has the controls (audit log, saved tests, human decisions) and the proof. Publishing them in the form reviewers need can shorten the longest part of every carrier deal.

Biggest Risk

Better-known claims vendors such as Hi Marley (13 of the top 30 P&C carriers) and Liberate (70+ carriers and brokers) already show carrier proof. A thin site with only workers' comp TPA stories can lose carrier deals before the demo, while a fast-growing team stretches forward-deployed implementation capacity.

01

Start here

The three moves to make first. Each one is written up in full below.

  1. 1

    Publish an NAIC AI Bulletin vendor pack on the Security page, so carrier AI reviews start with answers

    List the AI and security questions from the last three carrier or TPA reviews, and map each one to an Oversight feature.

    High impact First result in 1 month
  2. 2

    Swap the personal demo calendar for a short routed demo request that asks segment, line, claims system and volume

    Write the four questions and set up a team calendar with round-robin routing.

    Medium impact First result in 2 weeks
  3. 3

    Put line-specific proof on every solution page, starting with workers' comp on Property & Casualty

    Tag the three case studies by line and segment.

    High impact First result in 1 month

02

Your idea portfolio

22 ideas, grouped by when to run them. Open any idea for how to run it: the steps, who to approach, copy you can use, and how to measure it.

  • 7 ideas are marked Unexpected: plays you are unlikely to have tried.

    1 of these are things Programmable can build and run for you.

Every idea placed by the effort it takes against the impact it could have. Quick wins sit top left; select a number to open that idea. The filters below narrow the list by status, timing, category and what Programmable can build.

Effort vs impactEach number is an idea. Select one to open it.
Impact →
Quick wins · 3Big bets · 11Fill-ins · 5Reconsider · 3
Effort →

Showing 22 of 22 ideas

This week 6

Quick wins with a clear first action.

  • 01

    The Carriers page promises "Built to pass your AI review". The Oversight page already has what such a review asks for: an append-only action log, saved tests on 40 claims, releases the customer approves, and coverage decisions kept with adjusters. But the Security page and Trust Center cover only SOC 2, HIPAA, GDPR, CCPA and TCPA. Neither mentions the NAIC Model Bulletin on the Use of AI Systems by Insurers. By March 2025, 23 states plus D.C. had adopted it. It requires insurers to check their third-party AI vendors and to put audit rights into vendor contracts. A rival, Wisedocs, already cites the bulletin on its homepage. As Avallon moves from TPAs to carriers, each carrier's review team has to map Avallon's controls to the bulletin itself.

    High impactMedium effort First result in 1 month Co founder

    First action

    List the AI and security questions from the last three carrier or TPA reviews, and map each one to an Oversight feature.

    How to measure it

    Carrier security and AI review time

    Now
    –
    Target
    -30%
    By
    2 quarters

    Steps

    1. 1

      Collect the questions

      Pull every AI and security question from recent reviews, questionnaires and RFPs.

      Co-founder, GTM 2 days

    2. 2

      Map to the bulletin

      One row per bulletin expectation, with the feature and the evidence that supports it.

      Forward Deployed Engineer 3 days

    3. 3

      Legal check

      Insurance regulatory counsel reviews the mapping and the sample audit-rights clause.

      Outside counsel 1 week

    4. 4

      Design the pack

      Summary page, mapping table, and screenshots of the log, saved tests and release approval.

      Founding Product Designer 3 days

    5. 5

      Publish

      An AI governance section on /security, the pack in the Trust Center, and a link from the Carriers page.

      Product Engineer 1 day

    6. 6

      Use in deals

      Send the pack after every carrier first call and log the questions it still leaves open.

      Co-founder, GTM ongoing

    7. 7

      Keep it current

      Update the pack as more states adopt the bulletin or add their own rules.

      Co-founder, GTM quarterly

    First result
    1 month
    Cost
    Staff time plus a few hours of regulatory counsel
    Owner
    Co founder, GTM

    Smallest test

    Send a two-page mapping to the next carrier in procurement and time their review against the last one.

    Why it works

    Add an "AI governance" section to /security and a vendor pack to the Trust Center. The pack has four parts. First, a table that maps each bulletin expectation (governance, risk controls, testing, data and record retention, vendor diligence, audit rights) to the Avallon feature that supports it, with screenshots of the action log, saved tests and release approval. Second, a sample audit-rights clause for the carrier's counsel to start from. Third, a completed AI vendor questionnaire. Fourth, a one-page summary the claims leader can forward. Link the pack from the "Built to pass your AI review" section of the Carriers page, and send it before the first security call.

    Carrier deals slow down in procurement and model-risk review, not in the demo. Answering the reviewer's checklist before it arrives shortens the cycle and shows the product was built for regulated claims. TPAs can also use the same pack for their carrier clients' audits.

    Ready to use

    ChecklistAI governance vendor pack

    [ ] One-page summary for the claims leader
    [ ] Bulletin expectation to Avallon feature mapping table
    [ ] Screenshots: action log, saved tests, release approval, escalation rules
    [ ] Sample audit-rights clause (counsel-reviewed)
    [ ] Completed AI vendor questionnaire
    [ ] Data and record retention statement
    [ ] Link from /security and the Carriers page
    [ ] Date and per-state note

    Risks and how to handle them

    • Overclaiming compliance

      Describe how each feature supports the insurer's own obligations. The bulletin is addressed to insurers, so never say Avallon "is compliant" with it.

    • States differ

      Have counsel keep a short per-state note and date the pack.

    Evidence

    • Your websiteSecurity page and Trust Center: SOC 2 Type II, HIPAA, GDPR, CCPA and TCPA; Avallon "records what each agent read, how it classified the information, what action it took, and when it involved a person"; the SOC 2 report is available on request. Neither page mentions the NAIC AI Model Bulletin, state insurance AI rules or AI-governance documents.Source
    • Industry researchNAIC Model Bulletin on the Use of AI Systems by Insurers: insurers must keep a written program for responsible AI use covering governance, risk management and internal audit, carry out diligence on third-party AI systems and data, and put audit rights in vendor contracts; regulators may ask about that diligence. 23 states plus D.C. had adopted it by March 2025.Source
    • Your websiteSolution pages (Carriers, Life & Health, Property & Casualty, Specialty): each shows the same three case studies and one Athens quote; none names a customer in its own line, and none names a carrier customer; the life, auto, property and specialty examples are product mockups for a fictional "Greenpoint Claims". The Carriers page has a section headed "Built to pass your AI review".Source
    • Your websiteOversight page: "Agents prepare claims. Adjusters decide them."; "Every call and workflow scored" against the client's criteria; an append-only action log that can be queried by API or sent to a SIEM; changes "ship as a tested version your team approves", with saved tests on 40 claims; coverage decisions go to an adjuster, and attorney mentions and distress go to a person.Source
    • Competitor researchWisedocs, "Decision intelligence for complex claims" for carriers, TPAs and legal teams, with a quantified case study per customer type and a homepage note that 25 jurisdictions have adopted the NAIC Model Bulletin; OutRival, AI "Digital Workers" with an insurance vertical, promising first workers live in about two weeks.Source
    • News coverageAlleyWatch (November 2025): $4.6 million seed led by Frontline Ventures with Y Combinator, 1984, Liquid2 and Booom; TPAs are the first market, "over 42,000 of them across the US and Europe", with carriers next; most deals start with an implementation fee, then a fixed volume price, a flat commitment or pay per successful outcome; a live nationwide TPA partner with more than 400 adjusters and contracts in the US and Europe.Source
  • 02

    Every "Book a demo" button, on every one of the 24 pages, opens one co-founder's personal cal.com booking page. Nothing asks whether the visitor is a carrier or a TPA, which line they handle, which claims system they run or how many claims they handle, and nothing records which page sent them. The Forward Deployed Engineer owns implementation from the first demo, and the team plans to more than double in six months, so first calls start blind and inbound will soon outgrow one calendar.

    Medium impactLow effort First result in 2 weeks Co founder

    First action

    Write the four questions and set up a team calendar with round-robin routing.

    How to measure it

    Demos that reach a proof of value

    Now
    –
    Target
    +20%
    By
    90 days

    Steps

    1. 1

      Write the questions

      Segment, lines, claims system, volume; all optional except segment.

      Co-founder, GTM 1 hour

    2. 2

      Team calendar

      Replace the personal link with a routed team event.

      Chief of Staff 1 day

    3. 3

      Capture source

      Record page and campaign on every booking.

      Product Engineer 1 day

    4. 4

      Sync to CRM

      Answers become fields on the deal.

      Chief of Staff 1 day

    5. 5

      Prep template

      A one-screen brief for the FDE built from the answers.

      Forward Deployed Engineer 1 day

    6. 6

      Review monthly

      Bookings by page, segment and line.

      Co-founder, GTM monthly

    First result
    2 weeks
    Cost
    $0 to a low monthly scheduling fee
    Owner
    Co founder, GTM

    Smallest test

    Add the segment question only and watch booking rate for two weeks.

    Why it works

    Put four questions before the calendar: carrier or TPA, main lines, claims system, and monthly claim or call volume. Capture the source page and campaign. Route carriers and TPAs to the right owner through a team calendar, and send the answers into the CRM and the FDE's call notes, so the demo opens with the visitor's own system and workflow.

    Qualified, prepared first calls convert better. The answers also show which solution pages and lines actually create pipeline, which no one can see today.

    Risks and how to handle them

    • Fewer bookings because of the extra questions

      Keep it to four, make three optional, and compare booking rate for two weeks.

    Evidence

    • Your websiteHomepage: title "Avallon | AI agents for insurance claims"; H1 "Claims without Constraints"; seven named agents (Claim intaker, Case status receptionist, Third party outreacher, Case copilot, Coverage & liability copilot, Repair liaison, Subrogation specialist); headline stats "88% Of repetitive workflows handled end-to-end", "31% Faster cycle times" and "4.6/5 CSAT scores with agent interactions" with no stated method, period or sample; sample scores "Calls 96% pass" and "Claim intake 98%"; logos of Athens Administrators, Guardian Life, Penser and Johnston & Associates; no pricing; every "Book a demo" button opens one co-founder's personal cal.com booking page.Source
    • HiringForward Deployed Engineer (posted March 6, 2026, still open): owns customer implementation from first demo to production, builds integrations to claims platforms and CRMs, and tunes prompts, voice and agent behaviour; "Expect to spend 30-50% of your time on hands-on technical work."Source
    • HiringAvallon's job posts: a Y Combinator company that went "from zero to seven figures of ARR in twelve months" and is "growing 3X quarter over quarter"; about 18 people, planning to "more than double in the next 6 months"; mission: "the automation layer for the entire insurance industry", starting with claims. Four roles are open, all on-site in New York, including a Founding Recruiter (posted October 3, 2026) who will own hiring across engineering, product, go-to-market and operations.Source
  • 03

    The Life & Health, Property & Casualty, Specialty and Carriers pages all show the same three case studies and one Athens quote. Their line examples are mockups for a fictional "Greenpoint Claims". Yet the real proof is line-specific: Johnston & Associates is live in workers' comp, with general liability validated and auto liability in development. Athens serves injured workers and medical providers. JAG Source extracts medical reports for workers' comp ancillary services and healthcare billing.

    High impactLow effort First result in 1 month Co founder

    First action

    Tag the three case studies by line and segment.

    How to measure it

    Solution-page demo rate

    Now
    –
    Target
    +25%
    By
    60 days

    Steps

    1. 1

      Tag stories

      Line, segment and use case for each case study.

      Co-founder, GTM 1 hour

    2. 2

      Match pages

      Which story leads each solution page.

      Co-founder, GTM 1 day

    3. 3

      Write proof blocks

      One stat, one quote and one sentence of context per page.

      Co-founder, GTM 2 days

    4. 4

      Design-partner offer

      For lines with no live customer yet.

      Co-founder, GTM 1 day

    5. 5

      Build filters

      Line and segment filters on /case-studies.

      Product Engineer 2 days

    6. 6

      Measure

      Demo rate per solution page.

      Co-founder, GTM monthly

    First result
    1 month
    Cost
    Staff time only
    Owner
    Co founder, GTM

    Smallest test

    Swap the proof block on the Property & Casualty page this week.

    Why it works

    Lead each page with the proof that matches it. Property & Casualty gets Johnston & Associates (workers' comp intake, general liability next) and Athens (workers' comp calls). Life & Health gets JAG Source's medical-document extraction as the closest proof, labelled honestly. Where there is no live customer yet (Specialty, life claims), say so and invite design partners. Tag each case study with line and segment, and add those filters to /case-studies.

    Claims leaders look for a peer in their own line. Real, labelled proof beats generic mockups, and an honest design-partner offer turns a weakness into a reason to call.

    Risks and how to handle them

    • Pages look thin in some lines

      Use the design-partner offer instead of mockups presented as proof.

    Evidence

    • Your websiteSolution pages (Carriers, Life & Health, Property & Casualty, Specialty): each shows the same three case studies and one Athens quote; none names a customer in its own line, and none names a carrier customer; the life, auto, property and specialty examples are product mockups for a fictional "Greenpoint Claims". The Carriers page has a section headed "Built to pass your AI review".Source
    • Your websiteJohnston & Associates case study: thousands of First Report of Injury forms a year; the agent handles "90%" of intake start to finish with "99%+" first-pass accuracy and is "80%" faster per claim (30-40 minutes by hand down to a few minutes). It began as a proof of value on real forms in a test environment, works inside the existing claims system and translates a Spanish claim form into English for adjusters. Workers' comp is live, general liability is validated end to end and slated for production next, and auto liability is in development. The COO: "we've already got ten more things on the list."Source
    • Your websiteJAG Source case study (workers' comp ancillary services and healthcare billing): "54,000" medical reports a year, "99%+" extraction accuracy and "60%" less time; Avallon's agents "extend JAG Source's existing technology stack"; planned next: referrals extraction, CMS-1500 extraction, web portal integration and PDF splitting. The CIO: "If you're not doing this, you're already behind."Source
  • 04

    The homepage and every solution page lead with "88% Of repetitive workflows handled end-to-end", "31% Faster cycle times" and "4.6/5 CSAT", with no period, sample or definition. The case studies have exact, sourced numbers (2,000+ calls a month, 64% resolved, 90% of intake, 99%+ accuracy). Insurers now run formal diligence on AI vendors and may be asked about it by regulators. An unsourced headline number is exactly what a model-risk reviewer flags.

    Medium impactLow effort First result in 1 week Co founder

    First action

    Write down where each of the three numbers comes from.

    How to measure it

    Review questions about headline metrics

    Now
    –
    Target
    Zero repeat questions
    By
    90 days

    Steps

    1. 1

      Trace each number

      Source data, customers, workflows and period.

      Forward Deployed Engineer 1 day

    2. 2

      Write definitions

      Handled end-to-end, cycle time and CSAT method.

      Co-founder, GTM 1 day

    3. 3

      Replace the weak ones

      Swap any number that can't be sourced for a case-study number.

      Co-founder, GTM 1 hour

    4. 4

      Publish a methods page

      Linked from every stat block.

      Product Engineer 1 day

    5. 5

      Add to the pack

      The same definitions in the vendor pack.

      Co-founder, GTM 1 hour

    6. 6

      Refresh

      Recalculate the numbers each quarter with the date shown.

      Forward Deployed Engineer quarterly

    First result
    1 week
    Cost
    $0
    Owner
    Co founder, GTM

    Smallest test

    Add a one-line footnote under the homepage stats today.

    Why it works

    Add a short "How we measure" note under the stats, linked to a methods page. It should say which customers and workflows are included, the period, how "handled end-to-end" and "cycle time" are defined, and how CSAT is collected. Where a number cannot be sourced, replace it with a case-study number that can. Repeat the same definitions in the vendor pack.

    Sourced numbers survive procurement and get repeated in internal business cases. A method note costs a day and protects every claim on the site.

    Risks and how to handle them

    • Numbers fall once defined strictly

      A lower, defensible number wins more carrier deals than a high, unexplained one.

    Evidence

    • Your websiteHomepage: title "Avallon | AI agents for insurance claims"; H1 "Claims without Constraints"; seven named agents (Claim intaker, Case status receptionist, Third party outreacher, Case copilot, Coverage & liability copilot, Repair liaison, Subrogation specialist); headline stats "88% Of repetitive workflows handled end-to-end", "31% Faster cycle times" and "4.6/5 CSAT scores with agent interactions" with no stated method, period or sample; sample scores "Calls 96% pass" and "Claim intake 98%"; logos of Athens Administrators, Guardian Life, Penser and Johnston & Associates; no pricing; every "Book a demo" button opens one co-founder's personal cal.com booking page.Source
    • Industry researchNAIC Model Bulletin on the Use of AI Systems by Insurers: insurers must keep a written program for responsible AI use covering governance, risk management and internal audit, carry out diligence on third-party AI systems and data, and put audit rights in vendor contracts; regulators may ask about that diligence. 23 states plus D.C. had adopted it by March 2025.Source
    • Your websiteAthens Administrators case study: callers, especially injured workers and medical providers, phoned repeatedly about bill status; an AI inbound receptionist linked to the claims system handles "2,000+" calls a month, resolves "64%" with no human and scores "4.6/5" CSAT. The VP of IT: "Avallon has made it simple to integrate AI into our claims operations." No date and no PDF; the only action is "Book a demo".Source
  • 05

    The three case studies are strong: 64% of calls resolved, 90% of intake handled with 99%+ accuracy, 54,000 reports a year. But they exist only as web pages with no date, no PDF and only a "Book a demo" button. The site has no resources section at all. A claims leader who wants to convince IT, compliance and finance has nothing to attach.

    Medium impactLow effort First result in 2 weeks Founding Product Designer

    First action

    Lay out the Johnston & Associates one-pager first.

    How to measure it

    One-pager downloads

    Now
    –
    Target
    3 per active deal
    By
    60 days

    Steps

    1. 1

      Template

      One page, same structure for all three.

      Founding Product Designer 1 day

    2. 2

      Write

      Problem, agent, systems, three numbers, quote.

      Co-founder, GTM 1 day

    3. 3

      Customer approval

      Each customer approves its page.

      Co-founder, GTM 1 week

    4. 4

      Add date and security line

      Publication date, certifications and Trust Center link.

      Founding Product Designer 1 hour

    5. 5

      Publish

      Ungated download button on each case study.

      Product Engineer 2 hours

    6. 6

      Use in follow-up

      Attach the matching one-pager after every demo.

      Forward Deployed Engineer ongoing

    First result
    2 weeks
    Cost
    Design time only
    Owner
    Founding Product Designer

    Smallest test

    Send the Johnston & Associates one-pager after the next three demos.

    Why it works

    Design a one-page PDF per case study: the problem, what the agent does, the systems it works in, three numbers, the customer quote, and a security line (SOC 2 Type II, HIPAA, link to the Trust Center). Add a "Download one-pager" button next to "Book a demo" with no form, and use the PDFs in every follow-up email.

    Enterprise deals are won by the internal champion. Giving them a forwardable page helps the story reach the people who sign.

    Risks and how to handle them

    • Customers hesitate to approve a PDF

      Use only content already public on the web page.

    Evidence

    • Your websiteAthens Administrators case study: callers, especially injured workers and medical providers, phoned repeatedly about bill status; an AI inbound receptionist linked to the claims system handles "2,000+" calls a month, resolves "64%" with no human and scores "4.6/5" CSAT. The VP of IT: "Avallon has made it simple to integrate AI into our claims operations." No date and no PDF; the only action is "Book a demo".Source
    • Your websiteSitemap (24 URLs, checked November 9, 2026): 8 platform pages, 5 solution pages (Life & Health, Property & Casualty, Specialty, Carriers, TPAs), 3 case studies, Security, About, Careers and legal pages (Privacy, Terms, Imprint). No blog, resources, pricing, comparison, per-integration or ROI pages, and no newsletter sign-up.Source
    • Your websiteJohnston & Associates case study: thousands of First Report of Injury forms a year; the agent handles "90%" of intake start to finish with "99%+" first-pass accuracy and is "80%" faster per claim (30-40 minutes by hand down to a few minutes). It began as a proof of value on real forms in a test environment, works inside the existing claims system and translates a Spanish claim form into English for adjusters. Workers' comp is live, general liability is validated end to end and slated for production next, and auto liability is in development. The COO: "we've already got ten more things on the list."Source
  • 22

    Avallon plans to more than double its roughly 18 people in six months, and growth depends on hiring forward-deployed engineers fast enough to implement new customers. The Ashby job posts carry the strongest pitch: "from zero to seven figures of ARR in twelve months", "growing 3X quarter over quarter", Y Combinator and named investors. But the careers page candidates land on shows only four roles, "Competitive compensation" and no funding, growth, team or customer figures.

    Medium impactLow effort First result in 2 weeks Founding Recruiter

    First action

    Copy the company paragraph from the job posts onto the careers page.

    How to measure it

    Qualified applicants per role

    Now
    –
    Target
    +50%
    By
    60 days

    Steps

    1. 1

      Copy the pitch

      Traction, investors and mission from the posts.

      Founding Recruiter 1 hour

    2. 2

      Add customer results

      The three case-study headlines with links.

      Founding Recruiter 1 hour

    3. 3

      Show pay ranges

      The ranges already in each post.

      Founding Recruiter 1 hour

    4. 4

      Why now

      One line per role on its impact.

      Chief of Staff 2 hours

    5. 5

      Publish

      Update /careers.

      Product Engineer 2 hours

    6. 6

      Measure

      Applicants per role, before and after.

      Founding Recruiter monthly

    First result
    2 weeks
    Cost
    $0
    Owner
    Founding Recruiter

    Smallest test

    Add the traction sentence to the top of /careers today.

    Why it works

    Move the job-post pitch onto /careers: the traction numbers, the investors, the three customer results, salary ranges per role (already in the posts), and a short "what you would work on" section linking to the case studies. Add one line per role on why it matters now.

    Candidates for early-stage roles weigh momentum and customers above all, and every implementation hire made sooner lets Avallon take on another customer.

    Risks and how to handle them

    • Numbers go stale

      Date them and refresh each quarter.

    Evidence

    • Your websiteCareers page: four open roles, in-person in New York City, "Competitive compensation" with "strong salary and equity packages"; no funding, revenue growth, team size or customer figures.Source
    • HiringAvallon's job posts: a Y Combinator company that went "from zero to seven figures of ARR in twelve months" and is "growing 3X quarter over quarter"; about 18 people, planning to "more than double in the next 6 months"; mission: "the automation layer for the entire insurance industry", starting with claims. Four roles are open, all on-site in New York, including a Founding Recruiter (posted October 3, 2026) who will own hiring across engineering, product, go-to-market and operations.Source
    • HiringForward Deployed Engineer (posted March 6, 2026, still open): owns customer implementation from first demo to production, builds integrations to claims platforms and CRMs, and tunes prompts, voice and agent behaviour; "Expect to spend 30-50% of your time on hands-on technical work."Source

This quarter 11

The core plays for the next three months.

  • 06

    The integrations page shows logos for 13 claims systems, from Guidewire, Duck Creek and FINEOS to Origami Risk and FileHandler, but none links anywhere or explains what the agent reads and writes in that system. "Works inside our own systems" was the first question Johnston & Associates asked. The Founding Product Designer post says customers are replacing software they have used for twenty years. Buyers search by the system they run, and the site has no page that answers them.

    High impactMedium effort First result in 1 quarter Forward Deployed Engineer

    First action

    List which claims systems current customers run and pick the first three.

    How to measure it

    Integration-page demos

    Now
    –
    Target
    10 a quarter
    By
    2 quarters

    Steps

    1. 1

      Rank systems

      By live customers and pipeline.

      Forward Deployed Engineer 1 day

    2. 2

      Write the template

      Reads, writes, sync, setup time, security, example.

      Forward Deployed Engineer 2 days

    3. 3

      Check with partners

      Confirm trademark and naming rules for each vendor.

      Co-founder, GTM 1 week

    4. 4

      Build pages

      /platform/integrations/<system>, linked from the logos.

      Product Engineer 1 week

    5. 5

      Add telephony pages

      Genesys, Five9 and Amazon Connect.

      Forward Deployed Engineer 1 week

    6. 6

      Measure

      Search impressions and demos per page.

      Co-founder, GTM monthly

    First result
    1 quarter
    Cost
    Engineering and writing time
    Owner
    Forward Deployed Engineer

    Smallest test

    Publish one page for the system most customers run.

    Why it works

    Build one page per claims system, starting with the systems live customers use. Each page covers what the agent reads, what it writes back, how status sync works, typical setup time, the security model, and a customer example where one exists. Add a short page for each telephony platform too (Genesys, Five9, Amazon Connect), since call agents are a lead use case.

    Searches that name a claims system come from real evaluators. Integration detail also removes the first objection every claims leader has.

    Risks and how to handle them

    • Implying a formal partnership that does not exist

      Describe the technical integration only, unless a partner agreement is in place.

    Evidence

    • Your websiteIntegrations page: logos only for 13 claims systems (Guidewire, FINEOS, Insurity, Duck Creek, Origami Risk, Riskonnect, FileHandler, Sapiens, Majesco, Snapsheet, Five Sigma, EIS, CCC), plus CRMs, email and chat tools, telephony (Twilio, Genesys, NICE, Five9, Amazon Connect), data warehouses and Model Context Protocol; no page or detail for any single integration.Source
    • HiringFounding Product Designer (posted October 2, 2026): owns every customer-facing surface, including the claims workspace, review flows and executive reporting, and designs how agents show "what the agent did, why, what it read, and where it was unsure"; "Our customers are replacing software they have used for twenty years"; the largest insurers will run real claims operations through the interface "in the next few months".Source
    • Your websiteSitemap (24 URLs, checked November 9, 2026): 8 platform pages, 5 solution pages (Life & Health, Property & Casualty, Specialty, Carriers, TPAs), 3 case studies, Security, About, Careers and legal pages (Privacy, Terms, Imprint). No blog, resources, pricing, comparison, per-integration or ROI pages, and no newsletter sign-up.Source
  • 07

    Avallon lists 13 claims systems it works with, and carriers are the next market. Carriers buy through the ecosystems of the core systems they run. Liberate, a rival selling to 70+ carriers and brokers, leads with its Guidewire and Duck Creek integrations. A listing or referral relationship with the systems Avallon's customers already run puts it in front of their other customers.

    High impactMedium effort First result in 2 quarters Co founder

    First action

    Rank the claims systems by number of Avallon customers and open deals.

    How to measure it

    Opportunities from claims-system partners

    Now
    –
    Target
    5 a quarter
    By
    3 quarters

    Steps

    1. 1

      Rank systems

      Customers and open deals per claims system.

      Co-founder, GTM 1 day

    2. 2

      Check each programme

      Requirements, fees and timelines from each vendor's own partner page.

      Chief of Staff 1 week

    3. 3

      Apply

      Two or three programmes first.

      Co-founder, GTM 1 month

    4. 4

      Joint one-pager

      Live customer, integration scope, setup time.

      Forward Deployed Engineer 1 week

    5. 5

      Brief partner teams

      Which client problems to refer.

      Co-founder, GTM ongoing

    6. 6

      Track

      Partner-sourced pipeline in the CRM.

      Chief of Staff monthly

    First result
    2 quarters
    Cost
    Programme fees vary; staff time
    Owner
    Co founder, GTM

    Smallest test

    Ask one shared customer to introduce Avallon to their claims-system account team.

    Why it works

    Pick the two or three claims systems with the most Avallon customers and pipeline. Apply to their partner or marketplace programmes, and build a joint one-pager with each: a live customer, what the integration does, and setup time. Ask each system's customer-success team which of their TPA and carrier clients struggle with intake or call volume.

    Partner channels bring warm introductions to buyers who already trust the core system, and listing is a credibility signal in carrier procurement.

    Risks and how to handle them

    • Programmes are slow or costly

      Start with the system with the most shared customers and a light programme.

    Builds on

    Evidence

    • Your websiteIntegrations page: logos only for 13 claims systems (Guidewire, FINEOS, Insurity, Duck Creek, Origami Risk, Riskonnect, FileHandler, Sapiens, Majesco, Snapsheet, Five Sigma, EIS, CCC), plus CRMs, email and chat tools, telephony (Twilio, Genesys, NICE, Five9, Amazon Connect), data warehouses and Model Context Protocol; no page or detail for any single integration.Source
    • Competitor researchHi Marley, "The Trusted Communication Fabric for P&C Insurance", 130+ customers including 13 of the top 30 P&C carriers, with named carrier case studies and outcome ranges such as 35-50% fewer inbound calls; Liberate, "Insurance-native AI, built by insurance people", with 70+ carriers and brokers and integrations with Guidewire and Duck Creek.Source
    • News coverageAlleyWatch (November 2025): $4.6 million seed led by Frontline Ventures with Y Combinator, 1984, Liquid2 and Booom; TPAs are the first market, "over 42,000 of them across the US and Europe", with carriers next; most deals start with an implementation fee, then a fixed volume price, a flat commitment or pay per successful outcome; a live nationwide TPA partner with more than 400 adjusters and contracts in the US and Europe.Source
  • 08

    Avallon's pricing is unusual and buyer-friendly. Deals start with an implementation fee, then the customer chooses a fixed volume price, a flat commitment, or pay per successful outcome. None of this is on the site, which has no pricing page. Meanwhile, the case studies give the inputs for a cost model: intake cut from 30-40 minutes by hand to a few minutes, and 64% of calls resolved with no human.

    High impactMedium effort First result in 6 weeks Co founder

    First action

    Write the three models in one paragraph each and agree what can be public.

    How to measure it

    Pricing-page demo conversion

    Now
    –
    Target
    5% of pricing-page visitors
    By
    90 days

    Steps

    1. 1

      Agree what's public

      Models and examples, with no list prices.

      Co-founder, GTM 2 days

    2. 2

      Build the model

      Inputs from the case studies: minutes per task and resolution rate.

      Forward Deployed Engineer 3 days

    3. 3

      Design

      Three models, one calculator, one call to action.

      Founding Product Designer 3 days

    4. 4

      Build

      Page and calculator, linked in the main navigation.

      Product Engineer 1 week

    5. 5

      Pass to the demo

      Calculator results prefill the request.

      Product Engineer 1 day

    6. 6

      Review

      Which model buyers pick, by segment.

      Co-founder, GTM monthly

    First result
    6 weeks
    Cost
    Engineering time
    Owner
    Co founder, GTM

    Smallest test

    Add a "How pricing works" paragraph to the TPAs page.

    Why it works

    Publish a "How pricing works" page that explains the three models in plain words, without list prices, and says when each one fits. Add a calculator: monthly claims or calls, minutes per task today, and loaded adjuster cost, giving hours returned and the break-even under pay-per-outcome. Pass the answers to the demo request.

    Pay-per-outcome removes the budget objection, but only if buyers know it exists. The calculator also hands the champion a business case for their CFO.

    Risks and how to handle them

    • Competitors learn the model

      The model is already in public press coverage; the advantage is explaining it best.

    Evidence

    • News coverageAlleyWatch (November 2025): $4.6 million seed led by Frontline Ventures with Y Combinator, 1984, Liquid2 and Booom; TPAs are the first market, "over 42,000 of them across the US and Europe", with carriers next; most deals start with an implementation fee, then a fixed volume price, a flat commitment or pay per successful outcome; a live nationwide TPA partner with more than 400 adjusters and contracts in the US and Europe.Source
    • Your websiteJohnston & Associates case study: thousands of First Report of Injury forms a year; the agent handles "90%" of intake start to finish with "99%+" first-pass accuracy and is "80%" faster per claim (30-40 minutes by hand down to a few minutes). It began as a proof of value on real forms in a test environment, works inside the existing claims system and translates a Spanish claim form into English for adjusters. Workers' comp is live, general liability is validated end to end and slated for production next, and auto liability is in development. The COO: "we've already got ten more things on the list."Source
    • Your websiteTPAs page: "Take on more claims with the team you have", "Every client's instructions, followed and cited", "Ready for every audit, renewal, and RFP"; "service levels are tracked per client"; no pricing, ROI calculator or downloadable material; the only action is "Book a demo".Source
  • 09

    Johnston & Associates started with a proof of value on real forms in a test environment, and that is how its two doubts (complex claims, and its own systems) were answered. The Oversight page already has the machinery: saved tests on 40 claims, and every change shipping as a tested version the customer approves. The step that wins deals exists inside the product, but buyers never see it on the site.

    High impactMedium effort First result in 1 month Forward Deployed Engineer

    First action

    Write the scope, data requirements and timeline of a 40-claim proof of value.

    How to measure it

    Proof-of-value conversion

    Now
    –
    Target
    60%
    By
    2 quarters

    Steps

    1. 1

      Define the scope

      40 claims or one week of calls, one workflow.

      Forward Deployed Engineer 2 days

    2. 2

      Data terms

      A short data-handling agreement for test data.

      Outside counsel 1 week

    3. 3

      Results report

      Per-claim scores, accuracy, time saved, exceptions.

      Founding Product Designer 1 week

    4. 4

      Price it

      Fixed fee credited against implementation.

      Co-founder, GTM 1 day

    5. 5

      Publish the offer

      On solution pages and in the demo follow-up.

      Product Engineer 2 days

    6. 6

      Track

      Time to result and conversion per proof of value.

      Co-founder, GTM monthly

    First result
    1 month
    Cost
    Forward-deployed time per proof of value
    Owner
    Forward Deployed Engineer

    Smallest test

    Offer it to the next three qualified demos.

    Why it works

    Package it as a named offer: send 40 past claims (or a week of calls), and within two weeks get the agent's results on each one, scored against your own criteria, with accuracy, time saved and the exceptions it sent to a person. Use the same saved-test view the customer will use in production. Put the offer on every solution page and fold the implementation fee into it.

    A buyer's own claims are the most convincing demo. A fixed, small scope makes the first yes easy, and the test set becomes the customer's acceptance test at go-live.

    Risks and how to handle them

    • Proofs of value eat engineering time

      Fix the scope at one workflow and two weeks, and charge a fee.

    Evidence

    • Your websiteOversight page: "Agents prepare claims. Adjusters decide them."; "Every call and workflow scored" against the client's criteria; an append-only action log that can be queried by API or sent to a SIEM; changes "ship as a tested version your team approves", with saved tests on 40 claims; coverage decisions go to an adjuster, and attorney mentions and distress go to a person.Source
    • Your websiteJohnston & Associates case study: thousands of First Report of Injury forms a year; the agent handles "90%" of intake start to finish with "99%+" first-pass accuracy and is "80%" faster per claim (30-40 minutes by hand down to a few minutes). It began as a proof of value on real forms in a test environment, works inside the existing claims system and translates a Spanish claim form into English for adjusters. Workers' comp is live, general liability is validated end to end and slated for production next, and auto liability is in development. The COO: "we've already got ten more things on the list."Source
    • HiringForward Deployed Engineer (posted March 6, 2026, still open): owns customer implementation from first demo to production, builds integrations to claims platforms and CRMs, and tunes prompts, voice and agent behaviour; "Expect to spend 30-50% of your time on hands-on technical work."Source
  • 10

    The TPAs page promises "Ready for every audit, renewal, and RFP" and says service levels are tracked per client. TPAs live on winning and renewing carrier and employer clients, and those clients now ask how their administrators use AI. Avallon could hand each TPA customer ready-made answers that make it look better in its own bids. Today the page has nothing to download.

    High impactMedium effort First result in 1 quarter Co founder

    First action

    Ask two TPA customers for the AI questions in their latest client RFPs.

    How to measure it

    TPA prospects requesting the kit

    Now
    –
    Target
    30 a quarter
    By
    2 quarters

    Steps

    1. 1

      Collect RFP questions

      From customers' recent client bids.

      Co-founder, GTM 1 week

    2. 2

      Write answers

      In the TPA's voice, editable.

      Co-founder, GTM 1 week

    3. 3

      SLA report template

      Per-client service levels from the agent logs.

      Product Engineer 1 week

    4. 4

      Diagram

      Agent work and adjuster decisions.

      Founding Product Designer 2 days

    5. 5

      Ship to customers

      Walk each TPA customer through it.

      Forward Deployed Engineer 2 weeks

    6. 6

      Offer to prospects

      On the TPAs page and in outbound.

      Co-founder, GTM ongoing

    First result
    1 quarter
    Cost
    Staff time only
    Owner
    Co founder, GTM

    Smallest test

    Give the RFP answers to one TPA customer before its next bid.

    Why it works

    Write a TPA kit: standard RFP answers on AI use, oversight, data handling and escalation, written in the TPA's voice; a per-client service-level report template; a diagram of where the agent works and where adjusters decide; and the vendor pack from idea 1. Give it to every TPA customer, and offer it to prospects as the reason to talk.

    When Avallon helps a TPA win clients, Avallon grows with the TPA's book. It also turns the TPA into an advocate inside its carrier clients, the market Avallon wants next.

    Risks and how to handle them

    • TPAs over-promise to their clients

      Keep the answers factual and reviewed by Avallon.

    Builds on

    Evidence

    • Your websiteTPAs page: "Take on more claims with the team you have", "Every client's instructions, followed and cited", "Ready for every audit, renewal, and RFP"; "service levels are tracked per client"; no pricing, ROI calculator or downloadable material; the only action is "Book a demo".Source
    • Your websiteSecurity page and Trust Center: SOC 2 Type II, HIPAA, GDPR, CCPA and TCPA; Avallon "records what each agent read, how it classified the information, what action it took, and when it involved a person"; the SOC 2 report is available on request. Neither page mentions the NAIC AI Model Bulletin, state insurance AI rules or AI-governance documents.Source
    • News coverageAlleyWatch (November 2025): $4.6 million seed led by Frontline Ventures with Y Combinator, 1984, Liquid2 and Booom; TPAs are the first market, "over 42,000 of them across the US and Europe", with carriers next; most deals start with an implementation fee, then a fixed volume price, a flat commitment or pay per successful outcome; a live nationwide TPA partner with more than 400 adjusters and contracts in the US and Europe.Source
  • 11

    Avallon's first market is TPAs, which it counts at over 42,000 across the US and Europe. All three of its case studies are in workers' comp or next to it: injured-worker and provider calls at Athens, First Report of Injury intake at Johnston & Associates, and medical reports for workers' comp ancillary services at JAG Source. That is a sharp wedge. Outreach that leads with a peer in the same line beats generic "AI for claims".

    High impactMedium effort First result in 1 month Co founder

    First action

    Pull 200 workers' comp TPAs with their claims system and adjuster count.

    How to measure it

    Outbound meetings

    Now
    –
    Target
    15 a month
    By
    90 days

    Steps

    1. 1

      Build the list

      Workers' comp TPAs, size and claims system.

      Chief of Staff 1 week

    2. 2

      Write sequences

      One per use case, each with a customer number.

      Co-founder, GTM 3 days

    3. 3

      Warm intros

      Ask customers for two peer introductions each.

      Co-founder, GTM 2 weeks

    4. 4

      Send

      Founder-sent, 50 a week.

      Co-founder, GTM weekly

    5. 5

      Offer the proof of value

      As the call to action.

      Forward Deployed Engineer ongoing

    6. 6

      Review

      Reply and meeting rates by use case.

      Co-founder, GTM weekly

    First result
    1 month
    Cost
    Data tools and founder time
    Owner
    Co founder, GTM

    Smallest test

    Send the FROI-intake sequence to 50 TPAs.

    Build the prospect list. The list compiled from public sources, verified, deduplicated and delivered to your CRM. Usually live within days.

    Book a working session

    Why it works

    Build a list of workers' comp TPAs by size and claims system. Write three short sequences, one per use case (status calls, FROI intake, medical reports), each opening with the matching customer number and an offer of the 40-claim proof of value. Ask existing customers for warm introductions within their industry associations.

    Same line, same forms and same pain make the proof directly transferable, and a narrow list keeps a small team focused.

    Risks and how to handle them

    • Over-concentration in one line

      Use workers' comp as the wedge and add general liability as Johnston & Associates goes live.

    Builds on

    Evidence

    • News coverageAlleyWatch (November 2025): $4.6 million seed led by Frontline Ventures with Y Combinator, 1984, Liquid2 and Booom; TPAs are the first market, "over 42,000 of them across the US and Europe", with carriers next; most deals start with an implementation fee, then a fixed volume price, a flat commitment or pay per successful outcome; a live nationwide TPA partner with more than 400 adjusters and contracts in the US and Europe.Source
    • Your websiteAthens Administrators case study: callers, especially injured workers and medical providers, phoned repeatedly about bill status; an AI inbound receptionist linked to the claims system handles "2,000+" calls a month, resolves "64%" with no human and scores "4.6/5" CSAT. The VP of IT: "Avallon has made it simple to integrate AI into our claims operations." No date and no PDF; the only action is "Book a demo".Source
    • Your websiteJohnston & Associates case study: thousands of First Report of Injury forms a year; the agent handles "90%" of intake start to finish with "99%+" first-pass accuracy and is "80%" faster per claim (30-40 minutes by hand down to a few minutes). It began as a proof of value on real forms in a test environment, works inside the existing claims system and translates a Spanish claim form into English for adjusters. Workers' comp is live, general liability is validated end to end and slated for production next, and auto liability is in development. The COO: "we've already got ten more things on the list."Source
    • Your websiteJAG Source case study (workers' comp ancillary services and healthcare billing): "54,000" medical reports a year, "99%+" extraction accuracy and "60%" less time; Avallon's agents "extend JAG Source's existing technology stack"; planned next: referrals extraction, CMS-1500 extraction, web portal integration and PDF splitting. The CIO: "If you're not doing this, you're already behind."Source
  • 12

    Customers expand once the first agent works. Johnston & Associates' COO says "we've already got ten more things on the list", with general liability next and auto liability in development. JAG Source has four extractions planned. The homepage names seven agents, from Claim intaker to Subrogation specialist. Most customers run one or two of them today.

    High impactLow effort First result in 1 quarter Forward Deployed Engineer

    First action

    Fill in the map for the three published customers.

    How to measure it

    Net revenue retention

    Now
    –
    Target
    150%+
    By
    4 quarters

    Steps

    1. 1

      Build the grid

      Agents by lines for each customer.

      Forward Deployed Engineer 2 days

    2. 2

      Size the next steps

      Volume and value of the next three.

      Forward Deployed Engineer 1 week

    3. 3

      Price them

      Under each customer's existing model.

      Co-founder, GTM 2 days

    4. 4

      Quarterly review

      Results so far and the next step.

      Co-founder, GTM quarterly

    5. 5

      Plan capacity

      Match expansions to FDE time.

      Chief of Staff monthly

    6. 6

      Track

      Expansion revenue by agent type.

      Chief of Staff monthly

    First result
    1 quarter
    Cost
    Staff time only
    Owner
    Forward Deployed Engineer

    Smallest test

    Bring a three-step expansion plan to one customer's next review.

    Why it works

    For each customer, map which of the seven agents and which lines they use, and the next three likely steps with their volume and value. Review the map in a quarterly business review with the customer's operations and IT leads, and price each step under the customer's existing model.

    Selling to a happy customer is cheaper than winning a new one, and outcome-based pricing grows automatically as agents take on more work.

    Risks and how to handle them

    • Expansion outruns implementation capacity

      Sequence expansions against forward-deployed hiring.

    Evidence

    • Your websiteJohnston & Associates case study: thousands of First Report of Injury forms a year; the agent handles "90%" of intake start to finish with "99%+" first-pass accuracy and is "80%" faster per claim (30-40 minutes by hand down to a few minutes). It began as a proof of value on real forms in a test environment, works inside the existing claims system and translates a Spanish claim form into English for adjusters. Workers' comp is live, general liability is validated end to end and slated for production next, and auto liability is in development. The COO: "we've already got ten more things on the list."Source
    • Your websiteJAG Source case study (workers' comp ancillary services and healthcare billing): "54,000" medical reports a year, "99%+" extraction accuracy and "60%" less time; Avallon's agents "extend JAG Source's existing technology stack"; planned next: referrals extraction, CMS-1500 extraction, web portal integration and PDF splitting. The CIO: "If you're not doing this, you're already behind."Source
    • Your websiteHomepage: title "Avallon | AI agents for insurance claims"; H1 "Claims without Constraints"; seven named agents (Claim intaker, Case status receptionist, Third party outreacher, Case copilot, Coverage & liability copilot, Repair liaison, Subrogation specialist); headline stats "88% Of repetitive workflows handled end-to-end", "31% Faster cycle times" and "4.6/5 CSAT scores with agent interactions" with no stated method, period or sample; sample scores "Calls 96% pass" and "Claim intake 98%"; logos of Athens Administrators, Guardian Life, Penser and Johnston & Associates; no pricing; every "Book a demo" button opens one co-founder's personal cal.com booking page.Source
  • 13

    The Founding Product Designer will own executive reporting, and the Oversight page already scores every call and workflow against the client's criteria. Those scores answer the questions a COO asks at renewal: calls resolved, hours returned, accuracy and exceptions. Today they live inside the product, and public case studies are written by hand, one at a time.

    High impactMedium effort First result in 2 months Founding Product Designer

    First action

    Define the five numbers every report shows.

    How to measure it

    Gross revenue retention

    Now
    –
    Target
    95%+
    By
    4 quarters

    Steps

    1. 1

      Pick the numbers

      Five metrics with clear definitions.

      Founding Product Designer 2 days

    2. 2

      Design

      One page for an executive.

      Founding Product Designer 1 week

    3. 3

      Build

      From the scoring and action log.

      Product Engineer 2 weeks

    4. 4

      Send

      Monthly to each sponsor.

      Forward Deployed Engineer monthly

    5. 5

      Case-study trigger

      Ask when thresholds are passed.

      Co-founder, GTM ongoing

    6. 6

      Watch usage

      Opens and replies per report.

      Founding Product Designer monthly

    First result
    2 months
    Cost
    Engineering time
    Owner
    Founding Product Designer

    Smallest test

    Hand-build the report for one customer this month.

    Why it works

    Send an automatic monthly report to each customer's executive sponsor. It covers volume handled, share resolved without a person, accuracy, time returned to adjusters, CSAT, exceptions and the trend, all in one page. Add a one-click "share as a quote" option. Ask for a case study whenever a customer passes a set threshold.

    Sponsors who see value monthly renew and expand, and the same numbers feed new public proof without extra work.

    Risks and how to handle them

    • A bad month is visible

      Show the trend and the action being taken.

    Evidence

    • HiringFounding Product Designer (posted October 2, 2026): owns every customer-facing surface, including the claims workspace, review flows and executive reporting, and designs how agents show "what the agent did, why, what it read, and where it was unsure"; "Our customers are replacing software they have used for twenty years"; the largest insurers will run real claims operations through the interface "in the next few months".Source
    • Your websiteOversight page: "Agents prepare claims. Adjusters decide them."; "Every call and workflow scored" against the client's criteria; an append-only action log that can be queried by API or sent to a SIEM; changes "ship as a tested version your team approves", with saved tests on 40 claims; coverage decisions go to an adjuster, and attorney mentions and distress go to a person.Source
    • Your websiteAthens Administrators case study: callers, especially injured workers and medical providers, phoned repeatedly about bill status; an AI inbound receptionist linked to the claims system handles "2,000+" calls a month, resolves "64%" with no human and scores "4.6/5" CSAT. The VP of IT: "Avallon has made it simple to integrate AI into our claims operations." No date and no PDF; the only action is "Book a demo".Source
  • 14

    Buried in the Johnston & Associates case study, the agent translates a Spanish claim form into English for adjusters. Athens' agent answers injured workers and medical providers around the clock. Claims teams that get forms and calls in more than one language have to route them to the few bilingual staff, yet the site never mentions languages.

    Medium impactMedium effort First result in 1 month Product Engineer

    First action

    Test the agents on forms and calls in the five most common non-English languages customers receive.

    How to measure it

    Language-led opportunities

    Now
    –
    Target
    10% of new pipeline
    By
    2 quarters

    Steps

    1. 1

      Test languages

      Documents, calls, texts and email.

      Product Engineer 1 week

    2. 2

      Set the bar

      Accuracy threshold before claiming a language.

      Forward Deployed Engineer 2 days

    3. 3

      Write the section

      On two platform pages, with the J&A example.

      Co-founder, GTM 1 day

    4. 4

      Add to the request

      A languages field on the demo request.

      Product Engineer 1 hour

    5. 5

      Report

      Language mix in customer reports.

      Product Engineer 2 weeks

    6. 6

      Review

      Quality by language each month.

      Forward Deployed Engineer monthly

    First result
    1 month
    Cost
    Testing time
    Owner
    Product Engineer

    Smallest test

    Add the J&A translation example to the Document Intelligence page.

    Why it works

    Confirm which languages the agents handle today for documents, calls, texts and email. Add a "Languages" section to the Omnichannel Agents and Document Intelligence pages, with the J&A example. Add a languages field to the demo request and the proof of value, and report language mix in the executive report.

    It is a concrete, easy-to-demonstrate capability that removes a real routing bottleneck and sets Avallon apart from template-based tools.

    Risks and how to handle them

    • Quality varies by language

      Only list languages that pass the accuracy bar, and route the rest to a person.

    Evidence

    • Your websiteJohnston & Associates case study: thousands of First Report of Injury forms a year; the agent handles "90%" of intake start to finish with "99%+" first-pass accuracy and is "80%" faster per claim (30-40 minutes by hand down to a few minutes). It began as a proof of value on real forms in a test environment, works inside the existing claims system and translates a Spanish claim form into English for adjusters. Workers' comp is live, general liability is validated end to end and slated for production next, and auto liability is in development. The COO: "we've already got ten more things on the list."Source
    • Your websiteAthens Administrators case study: callers, especially injured workers and medical providers, phoned repeatedly about bill status; an AI inbound receptionist linked to the claims system handles "2,000+" calls a month, resolves "64%" with no human and scores "4.6/5" CSAT. The VP of IT: "Avallon has made it simple to integrate AI into our claims operations." No date and no PDF; the only action is "Book a demo".Source
  • 15

    Avallon's customers already say unusually strong things in public. JAG Source's CIO: "If you're not doing this, you're already behind." Johnston & Associates' COO has "ten more things on the list". Athens' VP of IT says Avallon made it "simple to integrate AI". Their roles (IT, operations) are the people carrier and TPA buyers most want to hear from. There is no reference programme on the site.

    High impactLow effort First result in 1 quarter Co founder

    First action

    Ask the three published customers' champions to join.

    How to measure it

    Referenced-deal win rate

    Now
    –
    Target
    +20 points
    By
    3 quarters

    Steps

    1. 1

      Design the programme

      Asks and benefits on one page.

      Co-founder, GTM 1 day

    2. 2

      Invite champions

      Start with the three published customers.

      Co-founder, GTM 2 weeks

    3. 3

      Reference rota

      Spread calls so no one is overused.

      Chief of Staff ongoing

    4. 4

      Advisory group

      First meeting on roadmap and governance.

      Co-founder, GTM 1 quarter

    5. 5

      Co-present

      One industry session with a customer.

      Co-founder, GTM 2 quarters

    6. 6

      Track

      References used against deal outcomes.

      Chief of Staff monthly

    First result
    1 quarter
    Cost
    Small events budget
    Owner
    Co founder, GTM

    Smallest test

    Book one reference call for the next carrier deal.

    Why it works

    Invite these and future champions to a small reference programme: a few reference calls a quarter, early access to new agents, a seat on a customer advisory group that meets twice a year, and co-presenting at an industry event. Track every reference call against deal outcome, and give advocates something useful in return, such as roadmap input.

    In a cautious, regulated buying process, peers in the same job are the most trusted proof. Advocates also hear about peers who are evaluating AI.

    Risks and how to handle them

    • Reference fatigue

      Cap calls per customer per quarter.

    Evidence

    • Your websiteJAG Source case study (workers' comp ancillary services and healthcare billing): "54,000" medical reports a year, "99%+" extraction accuracy and "60%" less time; Avallon's agents "extend JAG Source's existing technology stack"; planned next: referrals extraction, CMS-1500 extraction, web portal integration and PDF splitting. The CIO: "If you're not doing this, you're already behind."Source
    • Your websiteJohnston & Associates case study: thousands of First Report of Injury forms a year; the agent handles "90%" of intake start to finish with "99%+" first-pass accuracy and is "80%" faster per claim (30-40 minutes by hand down to a few minutes). It began as a proof of value on real forms in a test environment, works inside the existing claims system and translates a Spanish claim form into English for adjusters. Workers' comp is live, general liability is validated end to end and slated for production next, and auto liability is in development. The COO: "we've already got ten more things on the list."Source
    • Your websiteAthens Administrators case study: callers, especially injured workers and medical providers, phoned repeatedly about bill status; an AI inbound receptionist linked to the claims system handles "2,000+" calls a month, resolves "64%" with no human and scores "4.6/5" CSAT. The VP of IT: "Avallon has made it simple to integrate AI into our claims operations." No date and no PDF; the only action is "Book a demo".Source
  • 16

    Avallon has no blog, resources or newsletter; the site is 24 pages. Yet the product generates exactly the kind of material claims leaders share: every call and workflow is scored, and the homepage already shows sample scores such as "Calls 96% pass" and "Claim intake 98%". The company is growing 3X quarter over quarter with a story most claims leaders have not heard.

    Medium impactLow effort First result in 2 months Co founder

    First action

    List ten observations from current deployments that customers have agreed can be shared in anonymised form.

    How to measure it

    Social-sourced demos

    Now
    –
    Target
    5 a month
    By
    2 quarters

    Steps

    1. 1

      Agree the rules

      What can be shared, anonymised and aggregated.

      Co-founder, GTM 1 day

    2. 2

      Build a backlog

      Ten observations from deployments.

      Forward Deployed Engineer 1 week

    3. 3

      Post weekly

      One number, one lesson, one image.

      Co-founder, GTM weekly

    4. 4

      Monthly email

      Best posts to prospects and customers.

      Co-founder, GTM monthly

    5. 5

      Insights page

      A simple archive on the site.

      Product Engineer 2 days

    6. 6

      Track

      Source question on the demo request.

      Chief of Staff monthly

    First result
    2 months
    Cost
    Founder time
    Owner
    Co founder, GTM

    Smallest test

    Post three numbers over three weeks and track profile visits and demo requests.

    Why it works

    Each week, a founder posts one short, anonymised observation from live agent data or a deployment: what share of status calls are about payments, how long First Report of Injury forms take by hand, which exceptions agents send to people. Turn the best into a monthly email for prospects, and collect them on a simple /insights page.

    Specific numbers from real claims operations earn attention in a niche where most AI content is generic, and they build the founder's credibility with carrier buyers.

    Risks and how to handle them

    • Sharing customer data

      Aggregate across customers and get sign-off before posting.

    Evidence

    • Your websiteSitemap (24 URLs, checked November 9, 2026): 8 platform pages, 5 solution pages (Life & Health, Property & Casualty, Specialty, Carriers, TPAs), 3 case studies, Security, About, Careers and legal pages (Privacy, Terms, Imprint). No blog, resources, pricing, comparison, per-integration or ROI pages, and no newsletter sign-up.Source
    • Your websiteOversight page: "Agents prepare claims. Adjusters decide them."; "Every call and workflow scored" against the client's criteria; an append-only action log that can be queried by API or sent to a SIEM; changes "ship as a tested version your team approves", with saved tests on 40 claims; coverage decisions go to an adjuster, and attorney mentions and distress go to a person.Source
    • HiringAvallon's job posts: a Y Combinator company that went "from zero to seven figures of ARR in twelve months" and is "growing 3X quarter over quarter"; about 18 people, planning to "more than double in the next 6 months"; mission: "the automation layer for the entire insurance industry", starting with claims. Four roles are open, all on-site in New York, including a Founding Recruiter (posted October 3, 2026) who will own hiring across engineering, product, go-to-market and operations.Source

Big bets 3

Larger plays that can change the trajectory.

  • 17

    The Product Engineer post says Avallon is building public claimant chat interfaces and branded portals for large carrier customers. The Founding Product Designer post says the largest insurers will soon run real claims operations through the product. A claimant-facing portal is visible and low-risk: the agent answers status and document questions while adjusters still decide. That makes it an easier first purchase for a cautious carrier than back-office automation.

    High impactHigh effort First result in 2 quarters Product Engineer

    First action

    Define the minimum portal scope from the first carrier build.

    How to measure it

    Portal-led carrier contracts

    Now
    –
    Target
    3
    By
    4 quarters

    Steps

    1. 1

      Scope

      Status, uploads, missing information, handoff.

      Product Engineer 2 weeks

    2. 2

      Make it configurable

      Branding and rules per carrier without custom code.

      Product Engineer 1 quarter

    3. 3

      Design for claimants

      Accessible, mobile-first, clear handoffs.

      Founding Product Designer 1 month

    4. 4

      Price

      Per claim served, under the outcome model.

      Co-founder, GTM 2 weeks

    5. 5

      Launch page

      With deployment timeline and security pack.

      Co-founder, GTM 2 weeks

    6. 6

      Measure

      Calls avoided, satisfaction, expansion to other agents.

      Product Engineer monthly

    First result
    2 quarters
    Cost
    A quarter of engineering focus
    Owner
    Product Engineer

    Smallest test

    Offer the portal as a separate line item in the next carrier proposal.

    Why it works

    Package the claimant portal and chat as a standalone entry product with a fixed scope: status, document upload, missing-information requests and handoff to an adjuster. Give it its own page, a published deployment timeline and an outcome-based price per claim served. Every portal then becomes the path to intake, outreach and copilot agents behind it.

    A visible, contained first product shortens carrier procurement, gives policyholders a better experience the carrier can point to, and creates a natural expansion path.

    Risks and how to handle them

    • Becomes a custom build per carrier

      Configuration only; say no to one-off features.

    • Claimant-facing errors

      Keep coverage decisions with adjusters and escalate distress to a person, as the Oversight rules already do.

    Evidence

    • HiringProduct Engineer (posted October 22, 2026): builds public claimant chat interfaces and branded portals for large carrier customers, plus auditable claim context and review and approval screens for agent work; watches real user behaviour with Datadog RUM; "This is not a frontend-only role."Source
    • HiringFounding Product Designer (posted October 2, 2026): owns every customer-facing surface, including the claims workspace, review flows and executive reporting, and designs how agents show "what the agent did, why, what it read, and where it was unsure"; "Our customers are replacing software they have used for twenty years"; the largest insurers will run real claims operations through the interface "in the next few months".Source
    • News coverageAlleyWatch (November 2025): $4.6 million seed led by Frontline Ventures with Y Combinator, 1984, Liquid2 and Booom; TPAs are the first market, "over 42,000 of them across the US and Europe", with carriers next; most deals start with an implementation fee, then a fixed volume price, a flat commitment or pay per successful outcome; a live nationwide TPA partner with more than 400 adjusters and contracts in the US and Europe.Source
  • 18

    Avallon scores every call and workflow and logs every action, across workers' comp, healthcare billing and more. Rivals publish customer outcome ranges (Hi Marley: 35-50% fewer inbound calls) and quantified case studies (Wisedocs), but nobody publishes a neutral benchmark of how claims work actually breaks down: why people call, how long intake takes, which documents cause exceptions. Avallon's data could be that reference.

    High impactHigh effort First result in 2 quarters Co founder

    First action

    Get customer agreement to include anonymised aggregate data, and list the ten metrics.

    How to measure it

    Benchmark-sourced pipeline

    Now
    –
    Target
    20 qualified opportunities
    By
    2 quarters after launch

    Steps

    1. 1

      Data agreement

      Customer consent for anonymised aggregates.

      Co-founder, GTM 1 month

    2. 2

      Pick metrics

      Ten that claims leaders care about.

      Forward Deployed Engineer 1 week

    3. 3

      Survey

      Short survey of claims leaders.

      Co-founder, GTM 1 month

    4. 4

      Analyse

      By line and segment, with method notes.

      Product Engineer 1 month

    5. 5

      Design and publish

      Report, summary and private peer benchmark.

      Founding Product Designer 1 month

    6. 6

      Promote

      Trade press, events and customer co-authors.

      Co-founder, GTM 1 quarter

    First result
    2 quarters
    Cost
    Staff time plus design and launch budget
    Owner
    Co founder, GTM

    Smallest test

    Publish one chart on why claimants call, from aggregated data.

    Why it works

    Publish an annual "State of claims operations" report from aggregated, anonymised agent data plus a short survey of claims leaders. Cover call reasons, intake times, document quality, exception rates and after-hours demand by line. Release it with customer commentary, give participants a private benchmark against peers, and present it at industry events.

    A proprietary benchmark gets cited by press, analysts and buyers, and it brings in claims leaders who want their own comparison. That makes it a lead source and a positioning asset in one.

    Risks and how to handle them

    • Sample too small or skewed

      State the sample clearly and combine with the survey.

    Builds on

    Evidence

    • Your websiteOversight page: "Agents prepare claims. Adjusters decide them."; "Every call and workflow scored" against the client's criteria; an append-only action log that can be queried by API or sent to a SIEM; changes "ship as a tested version your team approves", with saved tests on 40 claims; coverage decisions go to an adjuster, and attorney mentions and distress go to a person.Source
    • Competitor researchHi Marley, "The Trusted Communication Fabric for P&C Insurance", 130+ customers including 13 of the top 30 P&C carriers, with named carrier case studies and outcome ranges such as 35-50% fewer inbound calls; Liberate, "Insurance-native AI, built by insurance people", with 70+ carriers and brokers and integrations with Guidewire and Duck Creek.Source
    • Competitor researchWisedocs, "Decision intelligence for complex claims" for carriers, TPAs and legal teams, with a quantified case study per customer type and a homepage note that 25 jurisdictions have adopted the NAIC Model Bulletin; OutRival, AI "Digital Workers" with an insurance vertical, promising first workers live in about two weeks.Source
  • 19

    Avallon counts its TPA market across the US and Europe (over 42,000), already has contracts with administrators in both, lists GDPR on its Security page and has an Imprint page. But every page, case study and integration is written for the US. There is no European path on the site.

    High impactHigh effort First result in 2 quarters Co founder

    First action

    Review the European contracts for the country with the most pipeline.

    How to measure it

    European TPA customers

    Now
    –
    Target
    5
    By
    4 quarters

    Steps

    1. 1

      Pick the country

      Existing customer, pipeline and language.

      Co-founder, GTM 2 weeks

    2. 2

      Check data residency

      Where data is processed and stored.

      Forward Deployed Engineer 1 month

    3. 3

      Local pages

      Language, systems, GDPR detail.

      Co-founder, GTM 1 month

    4. 4

      Local proof

      One approved customer story.

      Co-founder, GTM 1 quarter

    5. 5

      Regional owner

      Hire or assign one person.

      Founding Recruiter 1 quarter

    6. 6

      Review

      Decide on a second country after four customers.

      Co-founder, GTM quarterly

    First result
    2 quarters
    Cost
    One regional hire plus localisation
    Owner
    Co founder, GTM

    Smallest test

    Translate the TPAs page for the first country and run a small outbound test.

    Why it works

    Pick one European country where Avallon already has a customer. Build a localised page set: local language, the claims systems used there, data residency and GDPR detail, and a local customer story once approved. Run the same 40-claim proof of value. Hire or assign one person to own the region before adding a second country.

    Fewer AI claims vendors focus on European TPAs, an existing customer gives local proof, and a GDPR-first story fits a market where data protection decides deals.

    Risks and how to handle them

    • Splitting focus too early

      One country, one owner, and a clear stop rule.

    Builds on

    Evidence

    • News coverageAlleyWatch (November 2025): $4.6 million seed led by Frontline Ventures with Y Combinator, 1984, Liquid2 and Booom; TPAs are the first market, "over 42,000 of them across the US and Europe", with carriers next; most deals start with an implementation fee, then a fixed volume price, a flat commitment or pay per successful outcome; a live nationwide TPA partner with more than 400 adjusters and contracts in the US and Europe.Source
    • Your websiteAbout page: backed by "Frontline Ventures, Y Combinator, 1984 Ventures, Liquid 2, and Boom VC"; "the agentic system of action for insurance claims"; headquartered in New York City; the site also has an Imprint page.Source
    • Your websiteSecurity page and Trust Center: SOC 2 Type II, HIPAA, GDPR, CCPA and TCPA; Avallon "records what each agent read, how it classified the information, what action it took, and when it involved a person"; the SOC 2 report is available on request. Neither page mentions the NAIC AI Model Bulletin, state insurance AI rules or AI-governance documents.Source

Wildcards 2

Unexpected experiments, each with a cheap way to test it.

  • 20

    JAG Source is not a TPA or a carrier. It provides workers' comp ancillary services and healthcare billing, and Avallon's agents "extend JAG Source's existing technology stack" to process 54,000 medical reports a year. Firms like this (bill review, case management, medical record and IME providers) serve many TPAs and carriers at once. Each one that builds on Avallon carries the technology into its whole client base.

    Medium impactMedium effort First result in 2 quarters Co founder

    First action

    Ask JAG Source whether it would co-present as the first "powered by" partner.

    How to measure it

    Partner-processed documents

    Now
    –
    Target
    20% of document volume
    By
    4 quarters

    Steps

    1. 1

      Define the programme

      API, schemas, pricing, co-branding.

      Co-founder, GTM 2 weeks

    2. 2

      First partner

      JAG Source as launch partner, if willing.

      Co-founder, GTM 1 month

    3. 3

      Partner docs

      API and schema guide.

      Product Engineer 1 month

    4. 4

      Recruit

      Three service firms with overlapping clients.

      Co-founder, GTM 1 quarter

    5. 5

      Joint stories

      One case study per partner.

      Co-founder, GTM 2 quarters

    6. 6

      Track

      Volume and end clients reached.

      Chief of Staff monthly

    First result
    2 quarters
    Cost
    Partner support time
    Owner
    Co founder, GTM

    Smallest test

    Pitch the programme to one service firm a current customer works with.

    Why it works

    Create a partner programme for claims service providers: API access, extraction schemas for their document types, partner pricing on volume, co-branded "powered by Avallon" material, and a joint case study. Recruit three firms whose clients overlap with Avallon's target TPAs.

    One partner deal reaches many end clients, and service firms already sell into the claims teams Avallon wants.

    Risks and how to handle them

    • Partners compete with direct sales

      Define which segments each partner serves.

    Evidence

    • Your websiteJAG Source case study (workers' comp ancillary services and healthcare billing): "54,000" medical reports a year, "99%+" extraction accuracy and "60%" less time; Avallon's agents "extend JAG Source's existing technology stack"; planned next: referrals extraction, CMS-1500 extraction, web portal integration and PDF splitting. The CIO: "If you're not doing this, you're already behind."Source
    • News coverageAlleyWatch (November 2025): $4.6 million seed led by Frontline Ventures with Y Combinator, 1984, Liquid2 and Booom; TPAs are the first market, "over 42,000 of them across the US and Europe", with carriers next; most deals start with an implementation fee, then a fixed volume price, a flat commitment or pay per successful outcome; a live nationwide TPA partner with more than 400 adjusters and contracts in the US and Europe.Source
  • 21

    The homepage already shows sample scores ("Calls 96% pass", "Claim intake 98%"), and the Oversight page scores every call and workflow against the client's criteria. Those numbers sit next to static claims with no date. Hi Marley publishes outcome ranges with a caveat. No claims AI vendor shows its agents' quality as a live, dated, aggregated number.

    Medium impactMedium effort First result in 1 quarter Product Engineer

    First action

    Decide which three aggregated metrics are stable enough to publish monthly.

    How to measure it

    Scorecard views in deals

    Now
    –
    Target
    50% of active deals
    By
    2 quarters

    Steps

    1. 1

      Pick metrics

      Three stable, aggregated measures.

      Product Engineer 1 week

    2. 2

      Customer consent

      Aggregate use in contracts or by agreement.

      Co-founder, GTM 1 month

    3. 3

      Build the pipeline

      Monthly aggregate export from scoring data.

      Product Engineer 2 weeks

    4. 4

      Design

      Definitions, period and trend.

      Founding Product Designer 1 week

    5. 5

      Publish

      Linked from homepage stats and the vendor pack.

      Product Engineer 2 days

    6. 6

      Govern

      Rules for what happens when a number drops.

      Co-founder, GTM monthly

    First result
    1 quarter
    Cost
    Engineering time
    Owner
    Product Engineer

    Smallest test

    Add the month and the sample size to the homepage sample scores.

    Why it works

    Publish a /scorecard page that updates monthly from aggregated, anonymised scoring data: pass rate by workflow type, share escalated to a person, and median time to completion. Each number gets a definition and the period it covers. Link it from the homepage stats and the vendor pack.

    Live, dated quality numbers are a bold trust signal in a market wary of AI, and they hold the team to a public standard.

    Risks and how to handle them

    • A visible dip

      Show trends and explain changes; honesty builds trust with risk teams.

    Builds on

    Evidence

    • Your websiteHomepage: title "Avallon | AI agents for insurance claims"; H1 "Claims without Constraints"; seven named agents (Claim intaker, Case status receptionist, Third party outreacher, Case copilot, Coverage & liability copilot, Repair liaison, Subrogation specialist); headline stats "88% Of repetitive workflows handled end-to-end", "31% Faster cycle times" and "4.6/5 CSAT scores with agent interactions" with no stated method, period or sample; sample scores "Calls 96% pass" and "Claim intake 98%"; logos of Athens Administrators, Guardian Life, Penser and Johnston & Associates; no pricing; every "Book a demo" button opens one co-founder's personal cal.com booking page.Source
    • Your websiteOversight page: "Agents prepare claims. Adjusters decide them."; "Every call and workflow scored" against the client's criteria; an append-only action log that can be queried by API or sent to a SIEM; changes "ship as a tested version your team approves", with saved tests on 40 claims; coverage decisions go to an adjuster, and attorney mentions and distress go to a person.Source
    • Competitor researchHi Marley, "The Trusted Communication Fabric for P&C Insurance", 130+ customers including 13 of the top 30 P&C carriers, with named carrier case studies and outcome ranges such as 35-50% fewer inbound calls; Liberate, "Insurance-native AI, built by insurance people", with 70+ carriers and brokers and integrations with Guidewire and Duck Creek.Source

03

How Programmable can help

Where an idea in this playbook is something Programmable builds or runs, it's listed here with what we'd do. Everything else is yours to run.

Have it built

ProgrammableAgency

1 of the 22 ideas are things Programmable builds and runs for its customers: pages, outreach, prospect lists, booking flows and monitors. Pick one, or send the whole list.

Prospect lists1The list compiled from public sources, verified, deduplicated and delivered to your CRM.

A 30-minute session with the engineer who would build it, not a sales call. You leave with a plan whether or not you hire us.

Book a working session

Usually live within days.

Tools for the rest

ProgrammableInfo

Turn the research into a list you can act on

These ideas start with finding the right companies. Info finds them and the right person at each, researches every one and drafts the outreach, then asks before it sends.

  • Build this list in Info for idea #10 (opens programmable.info in a new tab)
  • Build this list in Info for idea #11 (opens programmable.info in a new tab)

New accounts: 100,000 free credits with code BETA100K

Find the companies (opens programmable.info in a new tab)
More example playbooks and reviews

More growth playbooks

Other companies we have researched, each read from what they are hiring for.

  • Home Landing growth strategyAn online shop for designer tea towels and table linen: printed and woven cotton and linen from ten European makers such as Le Jacquard…24 growth ideas
  • The Sleep Company growth strategyAn Indian comfort-tech brand selling SmartGRID mattresses, pillows, bedding, ergonomic office chairs, desks, recliner sofas, massagers and…23 growth ideas
  • Solink growth strategyA cloud platform that turns the cameras a business already has into AI "vision intelligence": cloud video management, 375+ data…22 growth ideas
  • OLAPLEX growth strategyPrestige hair care built on patented bond-building technology: treatments, shampoos, conditioners, masks, stylers and new lash, brow and…21 growth ideas
  • Avanai growth strategyA forward-deployed engineering firm for agentic AI, based in the Frankfurt area. Avanai builds sovereign, governed agent layers and…23 growth ideas
  • SPAN growth strategySmart electrical panels (MLO 48, MLO 24, MAIN 40, 32 and 16), the SPAN Edge at-the-meter device, the SPAN Drive EV charger and the SPAN…21 growth ideas
  • Kismet Mosaic growth strategyAn online mosaic supply store selling its own Kismet recycled glass tile line (made in Istanbul, 97% recycled glass) alongside ceramic…23 growth ideas
  • Spara growth strategyAn AI agent platform for go-to-market teams: agents that qualify inbound leads across chat, email, voice and text, run AI product demos…21 growth ideas

See every playbook and website review

Independent research. Programmable wrote this playbook from public sources, as they stood on November 10, 2026, and cites them throughout. Avallon did not commission, review or endorse it, and Programmable is not affiliated with Avallon. Company and product names are trademarks of their owners, used only to identify them. The ideas are suggestions and opinion, not professional, legal or financial advice, and not statements about how Avallon runs its business.

Something wrong, or do you represent Avallon? Email [email protected] and we will correct it or take it down.

30/60/90 roadmap

What to focus on in each month, and the ideas that carry it.

  1. Days 1 to 306 ideas

    Arm the buying committee: the NAIC vendor pack, sourced headline numbers, a routed demo request, line-specific proof, shareable case-study PDFs and a careers page that sells the traction.

  2. Days 31 to 609 ideas

    Build demand and deal flow: integration pages, partner programmes, pricing logic, the 40-claim proof of value, the TPA RFP kit, workers' comp outbound, multilingual handling, references and founder-led posts.

  3. Days 61 to 907 ideas

    Expand and commit to the bets: customer expansion maps, executive reports, the claimant portal, the annual benchmark, Europe, a powered-by programme and a live scorecard.

The research behind the plan

3 research areas, each summarised in a few lines. Open any of them for the full detail.

Customers

What customers say

Customers value agents that work inside the claims system they already run. The most cited problem: adjusters pulled off complex claims by status calls and data entry.

  • Measured results: 64% of calls resolved, 90% of intake handled, 99%+ accuracy
  • Unstandardised forms and medical reports that template tools cannot read
  • In their words: “claim intake”

Market

Competitive landscape

4 competitors compared: Hi Marley, Liberate, Wisedocs and OutRival.

  • Gap to exploit: Public, line-by-line proof for TPAs
  • Hi Marley: Communication-led rather than end-to-end claim work
  • Liberate: No TPA focus on its homepage

Hiring signals

Avallon is about 18 people, growing 3X quarter over quarter, and plans to more than double in six months.

  • Founding Recruiter: Team to more than double in six months. Idea 22.
  • Forward Deployed Engineer: Implementation from first demo to production. Ideas 2, 9, 12.
  • Founding Product Designer: Largest insurers going live; executive reporting. Ideas 6, 13.

From the team that wrote this playbook

All products
  • ProgrammableAgencyWorking AI systems in days, not quarters (opens programmable.agency in a new tab)
  • ProgrammableInfoAn agent that finds, researches and asks firstNew accounts get 100,000 free enrichment credits with code BETA100K (opens programmable.info in a new tab)
  • ProgrammableDigitalScreenshots in, an animated product demo out (opens programmable.digital in a new tab)
  • ProgrammableWatchKnow the day a supplier falls out of compliance (opens programmable.watch in a new tab)
  • ProgrammableTalkGet registrants for your events, and know who’s coming (opens programmable.talk in a new tab)
  • ProgrammableMediaAn AI note taker, interpreter and webinar host for your calls (opens programmable.media in a new tab)
  • ProgrammablePageText what you need, your customer gets a page that works (opens programmable.page in a new tab)
  • ProgrammableCommunityForums with labelled AI members, never empty (opens programmable.community in a new tab)
ProgrammableMarketing
ExamplesLeaderboardsProductsTermsPrivacyCookiesDPALegalContact

© 2026 AI-Powered Growth Intelligence