edisyl growth playbook

Build one use-case page around the example the homepage already uses (or whichever first use case won deals show is stronger): the question a commercial analyst asks, the wrong answer an agent gives without context, the right answer with a Knowledge Pack, and what was taught (the definitions, the regulatory codes, the source tables). Show the three steps from the docs (connect a source, teach, ask), the deployment model (your cloud or on premises) and one call to action. Link it from the homepage hero.

A buyer understands a broad platform through one problem they recognise. A concrete before-and-after gives sales and the channel partner one story to tell, and lets prospects judge fit without a call.

1. Headline: Your AI keeps picking the wrong definition. Here is the fix.
2. The question: "How many active patients did we add in the Northeast last quarter?"
3. Without a Knowledge Pack: a confident answer using the wrong definition
4. With a Knowledge Pack: the answer your commercial team would give, with the definition and source cited
5. What was taught: three definitions reconciled, regulatory codes, source tables
6. How it works: connect a source, teach, ask
7. Runs in your cloud or on premises, inside the AI tools you already use
8. Call to action: Tell us the question your AI gets wrong

• Example too narrow for other industries

  Present it as the first of several examples and add a second once a second use case wins.

• Customer data in the example

  Use invented but realistic data and definitions.

• Your websiteHomepage (checked October 22, 2026): "The knowledge layer for enterprise AI." A Knowledge Pack has two halves, domain knowledge ("partner data, government and regulatory sources, market data, standards and codes", "built and maintained by fleets of agents, steered by subject matter experts") and proprietary knowledge, over a semantic layer; installed "in your cloud or on your premises". The only concrete example: models "will still be guessing which of your three definitions of an active patient your commercial team uses." "Tell us what you are building and we will tell you whether the knowledge layer fits." There are no industry or use-case pages, customer names, case studies, demo video or pricing; the sitemap has two URLs (home and careers).Source
• HiringProduct Marketing Lead (Boston / Remote, posted September 18, 2026): "edisyl has built a product that works, is gaining traction, and can solve many different problems. That creates a marketing challenge." The role decides "The best first use case for each type of customer", "Why edisyl is different from existing options" and "How customers can expand their use of the product over time"; it will "Study successful and unsuccessful sales opportunities", "Build case studies and customer stories" and create "the website, sales deck, product pages, demos". Year-one success: "The right buyers can understand what edisyl does in a few sentences". "This is not mainly a demand generation or content production role."Source
• Your websiteDocumentation (docs.edisyl.com, updated October 20, 2026): a CLI and Knowledge Pack platform with concepts "Ask, Teach, Task", "Known, Understood, New", sources and the semantic layer; Getting Started covers "Install & log in", "Connect a data source", "Teach & ask"; guides cover "Lexicon and objectives, closing gaps, steering with PACK.md, scheduling recurring work"; packs can be asked from common chat and coding tools, and the docs themselves can be searched from an AI tool.Source
• Your websiteCareers page: the Product Marketing Lead is described as "Founding marketing partner, not a demand-gen hire", with the line "The product works. The hard part now is saying what it's for, in a few sentences, to the right person." The company has no blog, newsletter or customer stories, and its GitHub organization has no public repositories.Source

Define a starter offer: one Knowledge Pack for one domain and one team, delivered in a set number of weeks, with entry criteria, deliverables, success measures and a fixed fee range. Write the partner one-pager and statement of work template.

A packaged first engagement lowers the partner's risk in proposing edisyl and turns their client relationships into a repeatable pipeline.

• Fixed scope underprices complex clients

  Set entry criteria and a clear upgrade path.

• HiringEnterprise Solutions Strategist / GTM Commercial Lead (posted October 7, 2026): "Your first job is to take over deals already in flight and close them." The role decides "how Edisyl reaches clients through our key channel partner", works "with a sales engineer alongside", grows "each client across teams and domains", and will "Make Edisyl easy for the firm to sell" with "positioning, proof points and commercial model that give partners the confidence to bring Edisyl to their clients, and to carry smaller opportunities on their own". Signals: selling "through or alongside a large professional services firm (Deloitte, Accenture, KPMG, or equivalent)".Source
• Your websiteHomepage (checked October 22, 2026): "The knowledge layer for enterprise AI." A Knowledge Pack has two halves, domain knowledge ("partner data, government and regulatory sources, market data, standards and codes", "built and maintained by fleets of agents, steered by subject matter experts") and proprietary knowledge, over a semantic layer; installed "in your cloud or on your premises". The only concrete example: models "will still be guessing which of your three definitions of an active patient your commercial team uses." "Tell us what you are building and we will tell you whether the knowledge layer fits." There are no industry or use-case pages, customer names, case studies, demo video or pricing; the sitemap has two URLs (home and careers).Source
• HiringProduct Marketing Lead (Boston / Remote, posted September 18, 2026): "edisyl has built a product that works, is gaining traction, and can solve many different problems. That creates a marketing challenge." The role decides "The best first use case for each type of customer", "Why edisyl is different from existing options" and "How customers can expand their use of the product over time"; it will "Study successful and unsuccessful sales opportunities", "Build case studies and customer stories" and create "the website, sales deck, product pages, demos". Year-one success: "The right buyers can understand what edisyl does in a few sentences". "This is not mainly a demand generation or content production role."Source

Run a short interview with each open and recently lost opportunity: the trigger, the first question they wanted answered, who signed off, what confused them, and the alternative they considered. Tag the answers and pick the lead problem by count, not opinion.

Positioning from real buyers avoids the trap of listing everything the product can do.

• Prospects in late-stage deals are busy

  Keep it to 20 minutes and offer a summary in return.

• HiringProduct Marketing Lead (Boston / Remote, posted September 18, 2026): "edisyl has built a product that works, is gaining traction, and can solve many different problems. That creates a marketing challenge." The role decides "The best first use case for each type of customer", "Why edisyl is different from existing options" and "How customers can expand their use of the product over time"; it will "Study successful and unsuccessful sales opportunities", "Build case studies and customer stories" and create "the website, sales deck, product pages, demos". Year-one success: "The right buyers can understand what edisyl does in a few sentences". "This is not mainly a demand generation or content production role."Source
• HiringEnterprise Solutions Strategist / GTM Commercial Lead (posted October 7, 2026): "Your first job is to take over deals already in flight and close them." The role decides "how Edisyl reaches clients through our key channel partner", works "with a sales engineer alongside", grows "each client across teams and domains", and will "Make Edisyl easy for the firm to sell" with "positioning, proof points and commercial model that give partners the confidence to bring Edisyl to their clients, and to carry smaller opportunities on their own". Signals: selling "through or alongside a large professional services firm (Deloitte, Accenture, KPMG, or equivalent)".Source

Record a two-minute demo: connect a source, teach one definition, ask the same question before and after. Put it under the homepage headline and in the docs introduction.

Seeing the change in an answer is more persuasive than any description of a knowledge layer.

• Demo goes out of date

  Re-record each quarter.

• Your websiteHomepage (checked October 22, 2026): "The knowledge layer for enterprise AI." A Knowledge Pack has two halves, domain knowledge ("partner data, government and regulatory sources, market data, standards and codes", "built and maintained by fleets of agents, steered by subject matter experts") and proprietary knowledge, over a semantic layer; installed "in your cloud or on your premises". The only concrete example: models "will still be guessing which of your three definitions of an active patient your commercial team uses." "Tell us what you are building and we will tell you whether the knowledge layer fits." There are no industry or use-case pages, customer names, case studies, demo video or pricing; the sitemap has two URLs (home and careers).Source
• Your websiteDocumentation (docs.edisyl.com, updated October 20, 2026): a CLI and Knowledge Pack platform with concepts "Ask, Teach, Task", "Known, Understood, New", sources and the semantic layer; Getting Started covers "Install & log in", "Connect a data source", "Teach & ask"; guides cover "Lexicon and objectives, closing gaps, steering with PACK.md, scheduling recurring work"; packs can be asked from common chat and coding tools, and the docs themselves can be searched from an AI tool.Source

Add a "talk to an engineer" link and a "request a sandbox pack" option to the docs header and the end of Getting Started, and link the docs prominently from the homepage. Track which docs pages lead to requests.

Practitioners who read the docs are the warmest leads; a clear next step turns their reading into pipeline.

• Hobbyist requests

  Ask for company and use case on the request form.

• Your websiteDocumentation (docs.edisyl.com, updated October 20, 2026): a CLI and Knowledge Pack platform with concepts "Ask, Teach, Task", "Known, Understood, New", sources and the semantic layer; Getting Started covers "Install & log in", "Connect a data source", "Teach & ask"; guides cover "Lexicon and objectives, closing gaps, steering with PACK.md, scheduling recurring work"; packs can be asked from common chat and coding tools, and the docs themselves can be searched from an AI tool.Source
• Your websiteHomepage (checked October 22, 2026): "The knowledge layer for enterprise AI." A Knowledge Pack has two halves, domain knowledge ("partner data, government and regulatory sources, market data, standards and codes", "built and maintained by fleets of agents, steered by subject matter experts") and proprietary knowledge, over a semantic layer; installed "in your cloud or on your premises". The only concrete example: models "will still be guessing which of your three definitions of an active patient your commercial team uses." "Tell us what you are building and we will tell you whether the knowledge layer fits." There are no industry or use-case pages, customer names, case studies, demo video or pricing; the sitemap has two URLs (home and careers).Source
• Your websiteContact form on the homepage: first and last name, email, company, a job role list (C-Suite Executive, VP / Head of Department, Director, Manager, Data / AI Practitioner, IT / Enterprise Architecture, Knowledge Management, Operations, Other) and "What are you trying to do?". A HubSpot chat widget and analytics run on the page.Source

Self-host and subset the fonts with font-display swap, drop one font family, load the chat widget after interaction, and keep one analytics pipeline. Re-test the page.

A fast first paragraph keeps first-time visitors reading, and fewer tools are easier to maintain.

• Losing tracking

  Check reports before removing a tool.

• Google PageSpeedMobile lab test: performance 64, largest contentful paint 7.8 s, and the largest element is the homepage's opening paragraph ("AI tools fail on expert work..."); 409 KB of third-party code from 8 services (a tag manager 193 KB, HubSpot 100 KB, Google Fonts 81 KB across three font families, plus a second analytics pipeline); 249 KB of unused JavaScript; 550 ms of render-blocking requests. Real visitors (Chrome UX Report) see LCP 1,464 ms, INP 62 ms and CLS 0.01, so the site passes Core Web Vitals for current traffic.Source
• Your websiteContact form on the homepage: first and last name, email, company, a job role list (C-Suite Executive, VP / Head of Department, Director, Manager, Data / AI Practitioner, IT / Enterprise Architecture, Knowledge Management, Operations, Other) and "What are you trying to do?". A HubSpot chat widget and analytics run on the page.Source

Publish a fair comparison: what each approach covers (metric definitions, internal documents, external regulations and standards), who maintains it, where it runs, and when you would use both. Give sales the same table.

Buyers already have a mental category; placing edisyl next to it is faster than inventing a new one alone.

• Sounding hostile to partners' tools

  Show where edisyl works with them.

• Competitor researchAdjacent options, from their own homepages: AtScale, "the semantic layer for trusted analytics" ("One metric, defined once... served the same way to every AI, BI tool, and team"); dbt Semantic Layer ("Define metrics once and deliver consistent, governed insights across tools"); Glean, "Enterprise AI that Works | Agents, Assistant & Search". None of them sells external domain knowledge (regulations, standards, codes) maintained for the customer.See the research Source
• Your websiteHomepage (checked October 22, 2026): "The knowledge layer for enterprise AI." A Knowledge Pack has two halves, domain knowledge ("partner data, government and regulatory sources, market data, standards and codes", "built and maintained by fleets of agents, steered by subject matter experts") and proprietary knowledge, over a semantic layer; installed "in your cloud or on your premises". The only concrete example: models "will still be guessing which of your three definitions of an active patient your commercial team uses." "Tell us what you are building and we will tell you whether the knowledge layer fits." There are no industry or use-case pages, customer names, case studies, demo video or pricing; the sitemap has two URLs (home and careers).Source
• HiringProduct Marketing Lead (Boston / Remote, posted September 18, 2026): "edisyl has built a product that works, is gaining traction, and can solve many different problems. That creates a marketing challenge." The role decides "The best first use case for each type of customer", "Why edisyl is different from existing options" and "How customers can expand their use of the product over time"; it will "Study successful and unsuccessful sales opportunities", "Build case studies and customer stories" and create "the website, sales deck, product pages, demos". Year-one success: "The right buyers can understand what edisyl does in a few sentences". "This is not mainly a demand generation or content production role."Source

Pick two customers with a clear before and after (answers that were wrong, time spent checking, a workflow automated). Write short stories with one number each, anonymised if needed.

Enterprise buyers and partners need proof that someone like them has done it.

• Customers decline

  Use anonymised industry descriptions.

• HiringProduct Marketing Lead (Boston / Remote, posted September 18, 2026): "edisyl has built a product that works, is gaining traction, and can solve many different problems. That creates a marketing challenge." The role decides "The best first use case for each type of customer", "Why edisyl is different from existing options" and "How customers can expand their use of the product over time"; it will "Study successful and unsuccessful sales opportunities", "Build case studies and customer stories" and create "the website, sales deck, product pages, demos". Year-one success: "The right buyers can understand what edisyl does in a few sentences". "This is not mainly a demand generation or content production role."Source
• Your websiteHomepage (checked October 22, 2026): "The knowledge layer for enterprise AI." A Knowledge Pack has two halves, domain knowledge ("partner data, government and regulatory sources, market data, standards and codes", "built and maintained by fleets of agents, steered by subject matter experts") and proprietary knowledge, over a semantic layer; installed "in your cloud or on your premises". The only concrete example: models "will still be guessing which of your three definitions of an active patient your commercial team uses." "Tell us what you are building and we will tell you whether the knowledge layer fits." There are no industry or use-case pages, customer names, case studies, demo video or pricing; the sitemap has two URLs (home and careers).Source
• Your websiteCareers page: the Product Marketing Lead is described as "Founding marketing partner, not a demand-gen hire", with the line "The product works. The hard part now is saying what it's for, in a few sentences, to the right person." The company has no blog, newsletter or customer stories, and its GitHub organization has no public repositories.Source

Build a set of 100 expert questions in one domain, answer them with an agent with and without a pack, have a subject matter expert grade the answers, and publish the method and results.

A transparent test turns a belief into evidence that sales, partners and buyers can check.

• Results look cherry-picked

  Publish the questions and the grading rules.

• Your websiteHomepage (checked October 22, 2026): "The knowledge layer for enterprise AI." A Knowledge Pack has two halves, domain knowledge ("partner data, government and regulatory sources, market data, standards and codes", "built and maintained by fleets of agents, steered by subject matter experts") and proprietary knowledge, over a semantic layer; installed "in your cloud or on your premises". The only concrete example: models "will still be guessing which of your three definitions of an active patient your commercial team uses." "Tell us what you are building and we will tell you whether the knowledge layer fits." There are no industry or use-case pages, customer names, case studies, demo video or pricing; the sitemap has two URLs (home and careers).Source
• Your websiteDocumentation (docs.edisyl.com, updated October 20, 2026): a CLI and Knowledge Pack platform with concepts "Ask, Teach, Task", "Known, Understood, New", sources and the semantic layer; Getting Started covers "Install & log in", "Connect a data source", "Teach & ask"; guides cover "Lexicon and objectives, closing gaps, steering with PACK.md, scheduling recurring work"; packs can be asked from common chat and coding tools, and the docs themselves can be searched from an AI tool.Source
• HiringProduct Marketing Lead (Boston / Remote, posted September 18, 2026): "edisyl has built a product that works, is gaining traction, and can solve many different problems. That creates a marketing challenge." The role decides "The best first use case for each type of customer", "Why edisyl is different from existing options" and "How customers can expand their use of the product over time"; it will "Study successful and unsuccessful sales opportunities", "Build case studies and customer stories" and create "the website, sales deck, product pages, demos". Year-one success: "The right buyers can understand what edisyl does in a few sentences". "This is not mainly a demand generation or content production role."Source

Build a partner kit: a ten-slide deck, discovery questions, qualification criteria, the demo video, the starter offer and two stories. Train one practice and collect feedback.

Consultants promote what they can explain; a kit makes edisyl easy to bring up with clients.

• Partner misrepresents the product

  Train first and review early proposals.

• HiringEnterprise Solutions Strategist / GTM Commercial Lead (posted October 7, 2026): "Your first job is to take over deals already in flight and close them." The role decides "how Edisyl reaches clients through our key channel partner", works "with a sales engineer alongside", grows "each client across teams and domains", and will "Make Edisyl easy for the firm to sell" with "positioning, proof points and commercial model that give partners the confidence to bring Edisyl to their clients, and to carry smaller opportunities on their own". Signals: selling "through or alongside a large professional services firm (Deloitte, Accenture, KPMG, or equivalent)".Source
• HiringProduct Marketing Lead (Boston / Remote, posted September 18, 2026): "edisyl has built a product that works, is gaining traction, and can solve many different problems. That creates a marketing challenge." The role decides "The best first use case for each type of customer", "Why edisyl is different from existing options" and "How customers can expand their use of the product over time"; it will "Study successful and unsuccessful sales opportunities", "Build case studies and customer stories" and create "the website, sales deck, product pages, demos". Year-one success: "The right buyers can understand what edisyl does in a few sentences". "This is not mainly a demand generation or content production role."Source

Write the map: the first pack, the adjacent teams that reuse its definitions, and the next domains. Add an account plan template and review each customer against it every quarter.

Expansion is cheaper than new logos, and a map makes it a plan rather than luck.

• Pushing too early

  Expand after the first pack shows results.

• HiringEnterprise Solutions Strategist / GTM Commercial Lead (posted October 7, 2026): "Your first job is to take over deals already in flight and close them." The role decides "how Edisyl reaches clients through our key channel partner", works "with a sales engineer alongside", grows "each client across teams and domains", and will "Make Edisyl easy for the firm to sell" with "positioning, proof points and commercial model that give partners the confidence to bring Edisyl to their clients, and to carry smaller opportunities on their own". Signals: selling "through or alongside a large professional services firm (Deloitte, Accenture, KPMG, or equivalent)".Source
• HiringProduct Marketing Lead (Boston / Remote, posted September 18, 2026): "edisyl has built a product that works, is gaining traction, and can solve many different problems. That creates a marketing challenge." The role decides "The best first use case for each type of customer", "Why edisyl is different from existing options" and "How customers can expand their use of the product over time"; it will "Study successful and unsuccessful sales opportunities", "Build case studies and customer stories" and create "the website, sales deck, product pages, demos". Year-one success: "The right buyers can understand what edisyl does in a few sentences". "This is not mainly a demand generation or content production role."Source
• Your websiteHomepage (checked October 22, 2026): "The knowledge layer for enterprise AI." A Knowledge Pack has two halves, domain knowledge ("partner data, government and regulatory sources, market data, standards and codes", "built and maintained by fleets of agents, steered by subject matter experts") and proprietary knowledge, over a semantic layer; installed "in your cloud or on your premises". The only concrete example: models "will still be guessing which of your three definitions of an active patient your commercial team uses." "Tell us what you are building and we will tell you whether the knowledge layer fits." There are no industry or use-case pages, customer names, case studies, demo video or pricing; the sitemap has two URLs (home and careers).Source

Write three short pages for the roles that appear most in deals: what they are trying to fix, what edisyl changes for them, the proof, and the next step.

Buyers stay when they see their own problem described in their own words.

• Too many pages for a small team

  Start with one role.

• Your websiteContact form on the homepage: first and last name, email, company, a job role list (C-Suite Executive, VP / Head of Department, Director, Manager, Data / AI Practitioner, IT / Enterprise Architecture, Knowledge Management, Operations, Other) and "What are you trying to do?". A HubSpot chat widget and analytics run on the page.Source
• HiringProduct Marketing Lead (Boston / Remote, posted September 18, 2026): "edisyl has built a product that works, is gaining traction, and can solve many different problems. That creates a marketing challenge." The role decides "The best first use case for each type of customer", "Why edisyl is different from existing options" and "How customers can expand their use of the product over time"; it will "Study successful and unsuccessful sales opportunities", "Build case studies and customer stories" and create "the website, sales deck, product pages, demos". Year-one success: "The right buyers can understand what edisyl does in a few sentences". "This is not mainly a demand generation or content production role."Source
• Your websiteHomepage (checked October 22, 2026): "The knowledge layer for enterprise AI." A Knowledge Pack has two halves, domain knowledge ("partner data, government and regulatory sources, market data, standards and codes", "built and maintained by fleets of agents, steered by subject matter experts") and proprietary knowledge, over a semantic layer; installed "in your cloud or on your premises". The only concrete example: models "will still be guessing which of your three definitions of an active patient your commercial team uses." "Tell us what you are building and we will tell you whether the knowledge layer fits." There are no industry or use-case pages, customer names, case studies, demo video or pricing; the sitemap has two URLs (home and careers).Source

List the domains and sources that are ready (by industry and regulator), with update frequency and an example question for each. Each entry becomes a page that can rank for searches on that regulation or standard.

A catalogue shows what is ready on day one and brings searchers looking for help with specific regulations.

• Revealing too much

  List sources, not methods.

• Your websiteHomepage (checked October 22, 2026): "The knowledge layer for enterprise AI." A Knowledge Pack has two halves, domain knowledge ("partner data, government and regulatory sources, market data, standards and codes", "built and maintained by fleets of agents, steered by subject matter experts") and proprietary knowledge, over a semantic layer; installed "in your cloud or on your premises". The only concrete example: models "will still be guessing which of your three definitions of an active patient your commercial team uses." "Tell us what you are building and we will tell you whether the knowledge layer fits." There are no industry or use-case pages, customer names, case studies, demo video or pricing; the sitemap has two URLs (home and careers).Source
• Your websiteDocumentation (docs.edisyl.com, updated October 20, 2026): a CLI and Knowledge Pack platform with concepts "Ask, Teach, Task", "Known, Understood, New", sources and the semantic layer; Getting Started covers "Install & log in", "Connect a data source", "Teach & ask"; guides cover "Lexicon and objectives, closing gaps, steering with PACK.md, scheduling recurring work"; packs can be asked from common chat and coding tools, and the docs themselves can be searched from an AI tool.Source

Publish a page on architecture, where data stays, access control, logging, updates and certifications in progress, with a downloadable brief for security reviewers.

Security review is where enterprise AI deals slow down; answering early saves weeks.

• Claiming certifications not held

  List only what is complete and what is planned.

• Your websiteHomepage (checked October 22, 2026): "The knowledge layer for enterprise AI." A Knowledge Pack has two halves, domain knowledge ("partner data, government and regulatory sources, market data, standards and codes", "built and maintained by fleets of agents, steered by subject matter experts") and proprietary knowledge, over a semantic layer; installed "in your cloud or on your premises". The only concrete example: models "will still be guessing which of your three definitions of an active patient your commercial team uses." "Tell us what you are building and we will tell you whether the knowledge layer fits." There are no industry or use-case pages, customer names, case studies, demo video or pricing; the sitemap has two URLs (home and careers).Source
• HiringEnterprise Solutions Strategist / GTM Commercial Lead (posted October 7, 2026): "Your first job is to take over deals already in flight and close them." The role decides "how Edisyl reaches clients through our key channel partner", works "with a sales engineer alongside", grows "each client across teams and domains", and will "Make Edisyl easy for the firm to sell" with "positioning, proof points and commercial model that give partners the confidence to bring Edisyl to their clients, and to carry smaller opportunities on their own". Signals: selling "through or alongside a large professional services firm (Deloitte, Accenture, KPMG, or equivalent)".Source

Publish one essay a month from the CEO on LinkedIn and the site, each on one belief with an example. Partners can share them with clients.

A clear point of view draws the buyers who already agree, and gives partners something to forward.

• Too abstract

  Open each essay with a real example.

• Your websiteHomepage (checked October 22, 2026): "The knowledge layer for enterprise AI." A Knowledge Pack has two halves, domain knowledge ("partner data, government and regulatory sources, market data, standards and codes", "built and maintained by fleets of agents, steered by subject matter experts") and proprietary knowledge, over a semantic layer; installed "in your cloud or on your premises". The only concrete example: models "will still be guessing which of your three definitions of an active patient your commercial team uses." "Tell us what you are building and we will tell you whether the knowledge layer fits." There are no industry or use-case pages, customer names, case studies, demo video or pricing; the sitemap has two URLs (home and careers).Source
• HiringProduct Marketing Lead (Boston / Remote, posted September 18, 2026): "edisyl has built a product that works, is gaining traction, and can solve many different problems. That creates a marketing challenge." The role decides "The best first use case for each type of customer", "Why edisyl is different from existing options" and "How customers can expand their use of the product over time"; it will "Study successful and unsuccessful sales opportunities", "Build case studies and customer stories" and create "the website, sales deck, product pages, demos". Year-one success: "The right buyers can understand what edisyl does in a few sentences". "This is not mainly a demand generation or content production role."Source
• Your websiteedisyl.com/llms.txt (the summary AI assistants read): edisyl "grew out of Flipside Crypto, Inc., founded in 2017", which built blockchain data across "more than 40 blockchains, 700 million resolved wallet identities, and seven trillion-plus rows of curated on-chain data"; "In May 2026, Flipside sold its blockchain data business to SonarX and now operates exclusively as edisyl"; CEO and co-founder Dave Balter. Its careers section still lists a VP/Director of Finance and a Product Manager, which are not on the current job board.Source

Build a portability test: the same pack connected to several of the tools customers use, answering the same questions, with results published. Offer it as a standard check in every pilot.

Showing portability answers the lock-in fear and makes the pack an asset the customer owns.

• Tools behave differently

  Report the differences honestly.

• Your websiteHomepage (checked October 22, 2026): "The knowledge layer for enterprise AI." A Knowledge Pack has two halves, domain knowledge ("partner data, government and regulatory sources, market data, standards and codes", "built and maintained by fleets of agents, steered by subject matter experts") and proprietary knowledge, over a semantic layer; installed "in your cloud or on your premises". The only concrete example: models "will still be guessing which of your three definitions of an active patient your commercial team uses." "Tell us what you are building and we will tell you whether the knowledge layer fits." There are no industry or use-case pages, customer names, case studies, demo video or pricing; the sitemap has two URLs (home and careers).Source
• Your websiteDocumentation (docs.edisyl.com, updated October 20, 2026): a CLI and Knowledge Pack platform with concepts "Ask, Teach, Task", "Known, Understood, New", sources and the semantic layer; Getting Started covers "Install & log in", "Connect a data source", "Teach & ask"; guides cover "Lexicon and objectives, closing gaps, steering with PACK.md, scheduling recurring work"; packs can be asked from common chat and coding tools, and the docs themselves can be searched from an AI tool.Source

Once the starter offer works with the first partner, sign three boutique consultancies with the same kit and margins. Keep the large partner's accounts protected.

A second channel spreads risk and reaches buyers the first partner does not.

• Channel conflict

  Clear rules on account ownership.

• HiringEnterprise Solutions Strategist / GTM Commercial Lead (posted October 7, 2026): "Your first job is to take over deals already in flight and close them." The role decides "how Edisyl reaches clients through our key channel partner", works "with a sales engineer alongside", grows "each client across teams and domains", and will "Make Edisyl easy for the firm to sell" with "positioning, proof points and commercial model that give partners the confidence to bring Edisyl to their clients, and to carry smaller opportunities on their own". Signals: selling "through or alongside a large professional services firm (Deloitte, Accenture, KPMG, or equivalent)".Source
• HiringProduct Marketing Lead (Boston / Remote, posted September 18, 2026): "edisyl has built a product that works, is gaining traction, and can solve many different problems. That creates a marketing challenge." The role decides "The best first use case for each type of customer", "Why edisyl is different from existing options" and "How customers can expand their use of the product over time"; it will "Study successful and unsuccessful sales opportunities", "Build case studies and customer stories" and create "the website, sales deck, product pages, demos". Year-one success: "The right buyers can understand what edisyl does in a few sentences". "This is not mainly a demand generation or content production role."Source

Publish a yearly report from anonymous pilot data and the accuracy tests: common failure types (definitions, regulations, sources), by domain, with examples. Launch it with the channel partner.

Original data earns press and gives partners a reason to start conversations.

• Customer confidentiality

  Aggregate and remove identifying detail.

• Your websiteHomepage (checked October 22, 2026): "The knowledge layer for enterprise AI." A Knowledge Pack has two halves, domain knowledge ("partner data, government and regulatory sources, market data, standards and codes", "built and maintained by fleets of agents, steered by subject matter experts") and proprietary knowledge, over a semantic layer; installed "in your cloud or on your premises". The only concrete example: models "will still be guessing which of your three definitions of an active patient your commercial team uses." "Tell us what you are building and we will tell you whether the knowledge layer fits." There are no industry or use-case pages, customer names, case studies, demo video or pricing; the sitemap has two URLs (home and careers).Source
• HiringProduct Marketing Lead (Boston / Remote, posted September 18, 2026): "edisyl has built a product that works, is gaining traction, and can solve many different problems. That creates a marketing challenge." The role decides "The best first use case for each type of customer", "Why edisyl is different from existing options" and "How customers can expand their use of the product over time"; it will "Study successful and unsuccessful sales opportunities", "Build case studies and customer stories" and create "the website, sales deck, product pages, demos". Year-one success: "The right buyers can understand what edisyl does in a few sentences". "This is not mainly a demand generation or content production role."Source
• HiringEnterprise Solutions Strategist / GTM Commercial Lead (posted October 7, 2026): "Your first job is to take over deals already in flight and close them." The role decides "how Edisyl reaches clients through our key channel partner", works "with a sales engineer alongside", grows "each client across teams and domains", and will "Make Edisyl easy for the firm to sell" with "positioning, proof points and commercial model that give partners the confidence to bring Edisyl to their clients, and to carry smaller opportunities on their own". Signals: selling "through or alongside a large professional services firm (Deloitte, Accenture, KPMG, or equivalent)".Source

Recruit part-time experts per domain (former regulators, compliance leads, analysts), pay per review, and name them on catalogue pages with their credentials.

Named experts make the domain half of the product credible to skeptical specialists.

• Expert availability

  Keep two per domain.

• Your websiteHomepage (checked October 22, 2026): "The knowledge layer for enterprise AI." A Knowledge Pack has two halves, domain knowledge ("partner data, government and regulatory sources, market data, standards and codes", "built and maintained by fleets of agents, steered by subject matter experts") and proprietary knowledge, over a semantic layer; installed "in your cloud or on your premises". The only concrete example: models "will still be guessing which of your three definitions of an active patient your commercial team uses." "Tell us what you are building and we will tell you whether the knowledge layer fits." There are no industry or use-case pages, customer names, case studies, demo video or pricing; the sitemap has two URLs (home and careers).Source
• Your websiteedisyl.com/llms.txt (the summary AI assistants read): edisyl "grew out of Flipside Crypto, Inc., founded in 2017", which built blockchain data across "more than 40 blockchains, 700 million resolved wallet identities, and seven trillion-plus rows of curated on-chain data"; "In May 2026, Flipside sold its blockchain data business to SonarX and now operates exclusively as edisyl"; CEO and co-founder Dave Balter. Its careers section still lists a VP/Director of Finance and a Product Manager, which are not on the current job board.Source

Offer a simple worksheet or form: a team lists how each department defines five key terms, and gets a one-page report of conflicts and the questions an AI agent would get wrong. Follow up with a pilot offer.

A useful free check starts the conversation with the exact problem edisyl solves.

• Shallow results

  Keep it focused and offer a call for depth.

• Your websiteHomepage (checked October 22, 2026): "The knowledge layer for enterprise AI." A Knowledge Pack has two halves, domain knowledge ("partner data, government and regulatory sources, market data, standards and codes", "built and maintained by fleets of agents, steered by subject matter experts") and proprietary knowledge, over a semantic layer; installed "in your cloud or on your premises". The only concrete example: models "will still be guessing which of your three definitions of an active patient your commercial team uses." "Tell us what you are building and we will tell you whether the knowledge layer fits." There are no industry or use-case pages, customer names, case studies, demo video or pricing; the sitemap has two URLs (home and careers).Source
• HiringProduct Marketing Lead (Boston / Remote, posted September 18, 2026): "edisyl has built a product that works, is gaining traction, and can solve many different problems. That creates a marketing challenge." The role decides "The best first use case for each type of customer", "Why edisyl is different from existing options" and "How customers can expand their use of the product over time"; it will "Study successful and unsuccessful sales opportunities", "Build case studies and customer stories" and create "the website, sales deck, product pages, demos". Year-one success: "The right buyers can understand what edisyl does in a few sentences". "This is not mainly a demand generation or content production role."Source

Update llms.txt with the current roles, the lead use case and links to docs and stories; review it monthly. Publish a small public repository with a sample pack and the docs examples.

For a company that sells context to AI, its own context should be accurate, and a sample repository lets developers try it.

• Sample exposes internals

  Use invented data.

• Your websiteedisyl.com/llms.txt (the summary AI assistants read): edisyl "grew out of Flipside Crypto, Inc., founded in 2017", which built blockchain data across "more than 40 blockchains, 700 million resolved wallet identities, and seven trillion-plus rows of curated on-chain data"; "In May 2026, Flipside sold its blockchain data business to SonarX and now operates exclusively as edisyl"; CEO and co-founder Dave Balter. Its careers section still lists a VP/Director of Finance and a Product Manager, which are not on the current job board.Source
• Your websiteCareers page: the Product Marketing Lead is described as "Founding marketing partner, not a demand-gen hire", with the line "The product works. The hard part now is saying what it's for, in a few sentences, to the right person." The company has no blog, newsletter or customer stories, and its GitHub organization has no public repositories.Source
• Your websiteDocumentation (docs.edisyl.com, updated October 20, 2026): a CLI and Knowledge Pack platform with concepts "Ask, Teach, Task", "Known, Understood, New", sources and the semantic layer; Getting Started covers "Install & log in", "Connect a data source", "Teach & ask"; guides cover "Lexicon and objectives, closing gaps, steering with PACK.md, scheduling recurring work"; packs can be asked from common chat and coding tools, and the docs themselves can be searched from an AI tool.Source