ProgrammableMarketing/nucleoresearch.com (opens in a new tab)
PDF

Views

  • Plan
  • Roadmap
  • Research

On this page

  • The brief
  • Start here
  • Idea portfolio
  • Effort vs impact
  • This week · 8
  • This quarter · 11
  • Big bets · 3
  • Wildcards · 2
  • How we can help
Generated Nov 10, 2026 1 view
  1. Programmable
  2. Playbooks
  3. Nucleo
Nucleo logo

growth playbook · Nov 10, 2026

nucleoresearch.com (opens in a new tab)

Nucleo growth playbook

Diagnosis

Nucleo is a two-person Y Combinator (Fall 2025) company with real hospital partners (Stanford Hospital, Cedars-Sinai, UCI Health, Weill Cornell) and strong claims: segmentation 2,500x faster than manual and 98% agreement with experts. Its own website hides almost all of that. The homepage is one page, has no H1, shows rolling digits ("0 1 2 ...

9%") where the two proof numbers should be, and names none of its hospitals or investors in the text. A registered validation study of its body-composition software, Soma, still reads "Not yet recruiting" with completion dates in June 2026. The first month is about making the proof visible: numbers in text, named partners, a clear intended-use line, an updated registry record and a sample report.

The next two months turn the one-page site into use-case and evidence pages, open research and pharma revenue that does not wait for FDA clearance, and use the new SAM.gov registration for federal research funding. The big bets are a narrow first FDA submission, a distribution partner and body-composition endpoints for pharma trials.

Growth Score

6/10

Biggest Opportunity

Sell what can be sold today, research and pharma trial analysis, using proof that already exists but is invisible on the website, while the clinical product works toward FDA clearance.

Biggest Risk

With a team of two, Nucleo competes with funded players such as Mint Medical (trial reads with RECIST compliance checks) and Median Technologies (an imaging CRO with an FDA-cleared product), and "automated cancer diagnostics" wording before clearance can slow hospital and procurement reviews.

01

Start here

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

  1. 1

    Put "2,500x faster" and "98% agreement with experts" in your homepage text: today the page shows rolling digits "0 1 2 ... 9%" instead

    Replace the two counters with static text and add an H1 today.

    High impact First result in 2 weeks
  2. 2

    Add one intended-use line: who Nucleo is for today, and what is still on the way to FDA clearance

    Ask regulatory counsel for a one-sentence intended-use statement.

    Medium impact First result in 1 month
  3. 3

    Name the hospitals you already list on Y Combinator on your own homepage

    Email each partner contact asking how they agree to be named.

    High impact First result in 2 weeks

02

Your idea portfolio

24 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.

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

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 · 5Fill-ins · 9Reconsider · 7
Effort →

Showing 24 of 24 ideas

This week 8

Quick wins with a clear first action.

  • 01

    Nucleo's two strongest proof points, segmentation "2,500x faster" than manual and "98% agreement between Nucleo and medical experts", appear on its Y Combinator profile but not in the text of its own homepage. In the homepage's text, the metrics block between "Faster than manual segmentation." and "Agreement between Nucleo and experts." reads "0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9%": the animated counters are built from digit reels, so search engines, AI assistants, link previews and screen readers get a string of digits, not the claim. The page also has no H1, so the main heading search engines look for is missing. A radiologist who pastes the URL into a chat or a procurement team running a text check sees no number at all.

    High impactLow effort First result in 2 weeks Founder & CEO

    First action

    Replace the two counters with static text and add an H1 today.

    How to measure it

    Homepage-to-demo rate

    Now
    –
    Target
    +30%
    By
    60 days

    Steps

    1. 1

      Write the two claims

      Exact wording, the study each comes from, sample size and year.

      Founder & CTO 2 hours

    2. 2

      Replace the counters

      Static text with the final value; hide the digit reels from screen readers if the animation stays.

      Freelance web designer half a day

    3. 3

      Add an H1

      Promote the hero heading to H1 and keep headings in order below it.

      Freelance web designer 1 hour

    4. 4

      Add source lines

      One line under each number; link to the registered validation study.

      Founder & CEO 1 hour

    5. 5

      Check as a machine sees it

      Text-only fetch, a LinkedIn and Slack link preview, and a screen reader pass.

      Founder & CTO 1 hour

    6. 6

      Request indexing

      Resubmit the homepage in Google Search Console.

      Founder & CTO 15 minutes

    7. 7

      Track

      Demo bookings and search impressions for the claims before and after.

      Founder & CEO monthly

    First result
    2 weeks
    Cost
    A few hours of designer time
    Owner
    Founder & CEO

    Smallest test

    Change the counters to static text today and check the page with a text-only fetch.

    Why it works

    Rewrite the metrics block so each number is real text: "2,500x faster than manual segmentation" and "98% agreement with expert radiologists", each followed by a short source line ("Internal validation, n = ..., [year]"; link to the registered Soma study once results are posted). Keep the animation as decoration if wanted, but hide the digit reels from assistive technology and put the final value in the text. Turn "Automated cancer diagnostics" into a real H1 that carries the outcome. Then check the result with a text-only fetch and a link preview in Slack or LinkedIn.

    Clinical buyers trust numbers they can read and source. Putting the claims in text makes them visible to search engines, AI answers, screen readers and the people who forward the link inside a hospital, and the source line answers the first question a radiologist will ask.

    Ready to use

    ChecklistHomepage proof checklist

    [ ] "2,500x faster than manual segmentation" as text
    [ ] "98% agreement with expert radiologists" as text
    [ ] A source line under each (study, n, year)
    [ ] Digit reels hidden from screen readers
    [ ] One H1, then h2/h3 in order
    [ ] Named hospitals and Y Combinator in text (idea 2)
    [ ] Text-only fetch shows both numbers
    [ ] LinkedIn and Slack previews checked

    Risks and how to handle them

    • Claims read as unsupported once they are in plain text

      Add the source line and sample size; link the study record.

    • Clinical wording before FDA clearance

      Have regulatory counsel review the wording with the intended-use line (idea 3).

    Evidence

    • Your websiteHomepage (checked twice on November 10, 2026): title "Nucleo | Automated Cancer Diagnostics"; no H1 heading; between "Faster than manual segmentation." and "Agreement between Nucleo and experts." the page text reads "0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9%", and "2,500" and "98%" appear nowhere; the sections "We are backed by", "We work with the best hospitals in the US and worldwide" and "Our technology is powered by industry leaders" name no one in the text; the only call to action is a Cal.com "Book a demo" link.Source
    • Y Combinator profileFall 2025 batch, team of 2, San Francisco; "The first Agentic platform for Oncology"; segmentation "2,500x faster" than manual, "state-of-the-art performance in sarcopenia detection", "98% agreement between Nucleo and medical experts" in validation studies; aims to cut the 9-11 week diagnosis-to-treatment window to 10 days or less; hospital partners listed: Stanford Hospital, Cedars-Sinai, UCI Health and Weill Cornell.Source
    • PageSpeedMobile performance 76 (desktop 96); in the lab the largest element (the "Nucleo" text) appears at 7.1s on a phone; the page weighs 2.4 MB, 1.5 MB of it images (about 966 KB saveable) plus a 280 KB animation script from unicorn.studio; real Chrome visitors pass Core Web Vitals (p75 LCP 0.72s, INP 95 ms, CLS 0); accessibility 86 with low-contrast text, out-of-order headings and links without names.Source
  • 02

    Nucleo's Y Combinator profile names Stanford Hospital, Cedars-Sinai, UCI Health and Weill Cornell as hospital partners, but the homepage section "We work with the best hospitals in the US and worldwide" names none of them in its text, and "We are backed by" does not say Y Combinator. Rival Mint Medical lists its hospital users by name on its homepage.

    High impactLow effort First result in 2 weeks Founder & CEO

    First action

    Email each partner contact asking how they agree to be named.

    How to measure it

    Homepage-to-demo rate

    Now
    –
    Target
    +20%
    By
    60 days

    Steps

    1. 1

      Ask permission

      Each hospital's communications office, with the exact wording proposed.

      Founder & CEO 1 week

    2. 2

      Describe the relationship

      Research, pilot or clinical, accurately.

      Founder & CEO 1 hour

    3. 3

      Add names and logos

      Text names, logos with alt text.

      Freelance web designer 2 hours

    4. 4

      Name the backers

      "Backed by Y Combinator" in text.

      Freelance web designer 15 minutes

    5. 5

      Add a quote

      One clinician quote, approved.

      Founder & CEO 2 weeks

    6. 6

      Reuse in outreach

      Same names in demo emails and decks.

      Founder & CEO ongoing

    First result
    2 weeks
    Cost
    $0
    Owner
    Founder & CEO

    Smallest test

    Add "Backed by Y Combinator" today and the first approved hospital name this week.

    Why it works

    With each partner's written approval, name the hospitals in text (and logos with alt text), say what kind of work it is (research collaboration, pilot, clinical use), and write "Backed by Y Combinator" in the backed-by section. Add one quote from a clinician if a partner agrees.

    Named academic centres are the strongest trust signal a two-person medical AI company can show, and text names are read by search engines, AI assistants and procurement teams.

    Risks and how to handle them

    • A hospital declines logo use

      Use a text description such as "a top-ten US academic cancer centre" for that partner.

    Evidence

    • Y Combinator profileFall 2025 batch, team of 2, San Francisco; "The first Agentic platform for Oncology"; segmentation "2,500x faster" than manual, "state-of-the-art performance in sarcopenia detection", "98% agreement between Nucleo and medical experts" in validation studies; aims to cut the 9-11 week diagnosis-to-treatment window to 10 days or less; hospital partners listed: Stanford Hospital, Cedars-Sinai, UCI Health and Weill Cornell.Source
    • Your websiteHomepage (checked twice on November 10, 2026): title "Nucleo | Automated Cancer Diagnostics"; no H1 heading; between "Faster than manual segmentation." and "Agreement between Nucleo and experts." the page text reads "0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9%", and "2,500" and "98%" appear nowhere; the sections "We are backed by", "We work with the best hospitals in the US and worldwide" and "Our technology is powered by industry leaders" name no one in the text; the only call to action is a Cal.com "Book a demo" link.Source
    • Competitor researchMint Medical: mint Lesion for structured oncology reads with RECIST/iRECIST "automated compliance checks" for clinical trials, pharma and CROs; names hospital users such as The Royal Marsden, Stanford University and German university hospitals.Source
  • 03

    The homepage title is "Automated Cancer Diagnostics" with no intended-use or regulatory statement, while Nucleo's own job post says it is "building toward FDA-cleared clinical software" and serves researchers and pharma teams as well as hospitals. A hospital reviewer or a pharma buyer has to guess what the product can be used for today.

    Medium impactLow effort First result in 1 month Founder & CEO

    First action

    Ask regulatory counsel for a one-sentence intended-use statement.

    How to measure it

    Demo-to-pilot rate

    Now
    –
    Target
    +15%
    By
    90 days

    Steps

    1. 1

      Draft the line

      Today's use and the clearance path.

      Founder & CEO 1 hour

    2. 2

      Counsel review

      Regulatory wording check.

      Regulatory consultant 1 week

    3. 3

      Publish

      Hero, footer and booking page.

      Freelance web designer 1 hour

    4. 4

      Align outputs

      Same label on reports and exports.

      Founder & CTO 1 day

    5. 5

      Brief the team

      One answer to "is it cleared?".

      Founder & CEO 30 minutes

    6. 6

      Review quarterly

      Update as the regulatory path moves.

      AI Engineer (open role) quarterly

    First result
    1 month
    Cost
    A few hours of regulatory advice
    Owner
    Founder & CEO

    Smallest test

    Add the line to the demo booking page first.

    Why it works

    Add a short, plain line near the top and in the footer: what Nucleo is used for today (for example research and trial analysis), and that clinical use is on the path to FDA clearance. Have regulatory counsel write the exact wording. Mirror it in the demo booking page.

    A clear intended-use statement removes a common reason clinical and procurement reviews stall, and it lets research and pharma buyers say yes now instead of waiting.

    Risks and how to handle them

    • Sounds less ambitious

      Pair it with the outcome (time to treatment) and the named partners.

    Evidence

    • HiringAI Engineer (San Francisco or remote, $70K-$120K, listed on Y Combinator): Nucleo turns "routine CT and MRI imaging into quantitative insights", works with "top-tier hospitals in the US and Europe" plus researchers and pharma teams, and is "building toward FDA-cleared clinical software"; the role enforces regulatory policies and tracks sign-offs, travels for on-site integrations and onboards new users; on-prem integration experience is a plus.Source
    • Your websiteHomepage content: three use cases, "Body composition and sarcopenia assessment", "Tumor Lesion Sizing" and "Target vs Non-Target Lesion Classification" (RECIST); meta description "Advanced AI for oncological medical imaging. From scan to answer. In seconds, not hours."; no product name, no intended-use or regulatory statement, and no privacy, security or careers link; the footer links to Y Combinator, LinkedIn and X.Source
  • 04

    The homepage says "Automated Cancer Diagnostics" and "Advanced AI for oncological medical imaging"; the Y Combinator profile says "The first Agentic platform for Oncology"; the launch post's promise is cutting 9-11 weeks to treatment to 10 days or less. The homepage headline has not changed since at least November 2025, and the product itself, Soma in the trial registry, is not named on the site.

    Medium impactLow effort First result in 1 month Founder & CEO

    First action

    Write three candidate lines and test them on five clinicians.

    How to measure it

    Branded search impressions

    Now
    –
    Target
    +50%
    By
    90 days

    Steps

    1. 1

      List current descriptions

      Site, YC, LinkedIn, X, deck.

      Founder & CEO 1 hour

    2. 2

      Draft three lines

      Outcome first, product named.

      Founder & CEO 2 hours

    3. 3

      Test with clinicians

      Five short calls or messages.

      Founder & CEO 3 days

    4. 4

      Update everywhere

      Homepage, meta, profiles.

      Freelance web designer half a day

    5. 5

      Name the product

      Use the product name consistently.

      Founder & CTO 1 hour

    6. 6

      Check AI answers

      Ask assistants what Nucleo does, monthly.

      Founder & CEO monthly

    First result
    1 month
    Cost
    $0
    Owner
    Founder & CEO

    Smallest test

    Change the meta description today.

    Why it works

    Choose one line built on the outcome buyers care about (faster, consistent oncology measurements that shorten time to treatment), name the product, and use it on the homepage, meta description, Y Combinator profile, LinkedIn, X and the demo page.

    One consistent description is what search engines, AI assistants and people forwarding the link repeat. Three different ones dilute a small company's footprint.

    Risks and how to handle them

    • "Agentic" means little to clinicians

      Lead with the clinical result; keep the technical word for investors.

    Evidence

    • Your websiteHomepage content: three use cases, "Body composition and sarcopenia assessment", "Tumor Lesion Sizing" and "Target vs Non-Target Lesion Classification" (RECIST); meta description "Advanced AI for oncological medical imaging. From scan to answer. In seconds, not hours."; no product name, no intended-use or regulatory statement, and no privacy, security or careers link; the footer links to Y Combinator, LinkedIn and X.Source
    • Y Combinator profileFall 2025 batch, team of 2, San Francisco; "The first Agentic platform for Oncology"; segmentation "2,500x faster" than manual, "state-of-the-art performance in sarcopenia detection", "98% agreement between Nucleo and medical experts" in validation studies; aims to cut the 9-11 week diagnosis-to-treatment window to 10 days or less; hospital partners listed: Stanford Hospital, Cedars-Sinai, UCI Health and Weill Cornell.Source
    • Your websiteSite record: domain registered August 2024; the homepage headline "Nucleo | Automated Cancer Diagnostics" is unchanged since at least November 2025; the site links only LinkedIn and X, with no newsletter, YouTube channel or blog.Source
    • ClinicalTrials.govNCT07600866, sponsored by Nucleo Research, Inc.: "Validation of a Body-Composition Segmentation Software (Soma)" on 200 publicly available CT scans; primary outcome Dice vs a multi-rater radiologist reference (thresholds 0.90 for muscle, subcutaneous and visceral fat, 0.85 intramuscular fat); secondary outcomes include Cohen's kappa for sarcopenia by SMI (threshold 0.80); status still "Not yet recruiting" with estimated start May 31, 2026 and primary completion June 15, 2026, last posted May 22, 2026.Source
  • 05

    Nucleo registered a validation study of its body-composition software, Soma (NCT07600866), on 200 public CT scans, with clear pass thresholds. The record was last posted May 22, 2026 and still says "Not yet recruiting", with an estimated primary completion of June 15, 2026, five months ago. Clinical buyers and researchers who check the registry see a study that looks stalled.

    Medium impactLow effort First result in 2 weeks Founder & CTO

    First action

    Log in to the registry account and correct the status and dates.

    How to measure it

    Deals citing the validation

    Now
    –
    Target
    3
    By
    90 days

    Steps

    1. 1

      Update status

      Actual start and completion dates.

      Founder & CTO 1 hour

    2. 2

      Summarise results

      Dice per tissue, kappa for sarcopenia, against the thresholds.

      Founder & CTO 1 week

    3. 3

      Post results

      Results section on the record.

      Founder & CTO 1 day

    4. 4

      Link it

      From the homepage numbers (idea 1).

      Freelance web designer 15 minutes

    5. 5

      Announce

      LinkedIn and X post with the headline result.

      Founder & CEO 1 hour

    6. 6

      Keep current

      Review every registered study quarterly.

      AI Engineer (open role) quarterly

    First result
    2 weeks
    Cost
    $0
    Owner
    Founder & CTO

    Smallest test

    Correct the status today.

    Why it works

    Update the record's status and actual dates, and post the results summary when ready. Then link the record from the homepage and every deck as Nucleo's public validation.

    A registered, pre-specified validation with posted results is rare among early medical AI companies and is exactly the evidence radiologists and pharma buyers look for.

    Risks and how to handle them

    • A tissue class misses its threshold

      Report it honestly with the next version's plan; transparency builds trust.

    Evidence

    • ClinicalTrials.govNCT07600866, sponsored by Nucleo Research, Inc.: "Validation of a Body-Composition Segmentation Software (Soma)" on 200 publicly available CT scans; primary outcome Dice vs a multi-rater radiologist reference (thresholds 0.90 for muscle, subcutaneous and visceral fat, 0.85 intramuscular fat); secondary outcomes include Cohen's kappa for sarcopenia by SMI (threshold 0.80); status still "Not yet recruiting" with estimated start May 31, 2026 and primary completion June 15, 2026, last posted May 22, 2026.Source
  • 06

    The homepage section "This is Nucleo in action" has no readable demo content, and the only next step is a Cal.com "Book a demo" link. Nucleo already validates Soma on publicly available CT scans, so it can show a real output without any patient data.

    High impactLow effort First result in 1 month Founder & CTO

    First action

    Pick one public abdominal CT from the validation cohort.

    How to measure it

    Sample-report-to-demo rate

    Now
    –
    Target
    10%
    By
    60 days

    Steps

    1. 1

      Pick scans

      Public, licence-checked CTs.

      Founder & CTO 1 day

    2. 2

      Run the tools

      Body composition, lesion sizing, RECIST.

      Founder & CTO 1 day

    3. 3

      Lay out the report

      Images, numbers, short explanations.

      Freelance web designer 2 days

    4. 4

      Publish

      Page plus PDF, labelled with intended use.

      Freelance web designer 1 day

    5. 5

      Add the button

      Next to Book a demo.

      Freelance web designer 1 hour

    6. 6

      Measure

      Views, downloads, demos after.

      Founder & CEO monthly

    First result
    1 month
    Cost
    A week of founder time
    Owner
    Founder & CTO

    Smallest test

    Post one annotated slice on LinkedIn with the numbers.

    Why it works

    Run Soma and the lesion tools on one or two public CT scans, publish the resulting report (segmentation images, muscle and fat areas, SMI, lesion sizes and RECIST classes) as a page and a downloadable PDF, and put "See a sample report" next to "Book a demo".

    Radiologists judge segmentation by eye. A real report answers "what do I get?" without a call, and gives the demo a head start.

    Risks and how to handle them

    • Dataset licence limits reuse

      Use scans whose licence allows publication and credit the source.

    Evidence

    • Your websiteHomepage (checked twice on November 10, 2026): title "Nucleo | Automated Cancer Diagnostics"; no H1 heading; between "Faster than manual segmentation." and "Agreement between Nucleo and experts." the page text reads "0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9%", and "2,500" and "98%" appear nowhere; the sections "We are backed by", "We work with the best hospitals in the US and worldwide" and "Our technology is powered by industry leaders" name no one in the text; the only call to action is a Cal.com "Book a demo" link.Source
    • ClinicalTrials.govNCT07600866, sponsored by Nucleo Research, Inc.: "Validation of a Body-Composition Segmentation Software (Soma)" on 200 publicly available CT scans; primary outcome Dice vs a multi-rater radiologist reference (thresholds 0.90 for muscle, subcutaneous and visceral fat, 0.85 intramuscular fat); secondary outcomes include Cohen's kappa for sarcopenia by SMI (threshold 0.80); status still "Not yet recruiting" with estimated start May 31, 2026 and primary completion June 15, 2026, last posted May 22, 2026.Source
  • 07

    Real Chrome visitors already pass Core Web Vitals (p75 LCP 0.72s), so this is not urgent. But in Google's phone test the main content takes 7.1s, the page weighs 2.4 MB, 1.5 MB of it images with about 966 KB saveable, and a 280 KB animation script loads from a third-party host. First-time visitors on a phone or a slow hospital guest network get the slow version.

    Low impactLow effort First result in 1 week Freelance web designer

    First action

    Re-export the three largest images.

    How to measure it

    Mobile LCP (lab)

    Now
    –
    Target
    under 3s
    By
    30 days

    Steps

    1. 1

      List heavy assets

      From the PageSpeed report.

      Freelance web designer 1 hour

    2. 2

      Re-export images

      Display size, WebP/AVIF.

      Freelance web designer 2 hours

    3. 3

      Lazy-load

      Below-the-fold images.

      Freelance web designer 1 hour

    4. 4

      Decide on the animation

      Keep, defer or replace with a video poster.

      Founder & CEO 30 minutes

    5. 5

      Retest

      PageSpeed mobile and desktop.

      Freelance web designer 30 minutes

    6. 6

      Watch field data

      Keep passing Core Web Vitals.

      Founder & CTO monthly

    First result
    1 week
    Cost
    Half a day of designer time
    Owner
    Freelance web designer

    Smallest test

    Compress the hero image and retest.

    Why it works

    Export images at display size in modern formats, lazy-load below-the-fold images, and decide whether the animation is worth 280 KB; if it stays, load it after the main content.

    A lighter page is faster for the first visit from a conference badge scan or a forwarded link, which is when a small company gets judged.

    Risks and how to handle them

    • Visual quality drops

      Compare side by side before publishing.

    Evidence

    • PageSpeedMobile performance 76 (desktop 96); in the lab the largest element (the "Nucleo" text) appears at 7.1s on a phone; the page weighs 2.4 MB, 1.5 MB of it images (about 966 KB saveable) plus a 280 KB animation script from unicorn.studio; real Chrome visitors pass Core Web Vitals (p75 LCP 0.72s, INP 95 ms, CLS 0); accessibility 86 with low-contrast text, out-of-order headings and links without names.Source
  • 08

    PageSpeed scores the homepage 86 for accessibility, with low-contrast text, headings out of order and links without names. Hospitals and universities often ask vendors about accessibility, and a missing H1 plus unnamed links is an easy finding for a reviewer.

    Low impactLow effort First result in 1 week Freelance web designer

    First action

    Name the icon links and fix the contrast colours.

    How to measure it

    PageSpeed accessibility

    Now
    –
    Target
    95+
    By
    30 days

    Steps

    1. 1

      Fix contrast

      WCAG AA colours.

      Freelance web designer 1 hour

    2. 2

      Order headings

      H1, then h2, h3.

      Freelance web designer 1 hour

    3. 3

      Name links

      Labels for icon and logo links.

      Freelance web designer 30 minutes

    4. 4

      Retest

      PageSpeed and a screen reader.

      Freelance web designer 30 minutes

    5. 5

      Write a note

      One-page accessibility statement for questionnaires.

      Founder & CEO 1 hour

    6. 6

      Repeat

      On every new page.

      Freelance web designer ongoing

    First result
    1 week
    Cost
    A few hours
    Owner
    Freelance web designer

    Smallest test

    Name the links today.

    Why it works

    Raise text contrast to WCAG AA, put headings in order under the new H1 (idea 1), and give every link and icon a readable name. Keep a one-page accessibility note for procurement questionnaires.

    It removes small objections from slow, formal buying processes and helps screen-reader users and crawlers alike.

    Risks and how to handle them

    • Design changes

      Adjust shades, not the layout.

    Builds on

    Evidence

    • PageSpeedMobile performance 76 (desktop 96); in the lab the largest element (the "Nucleo" text) appears at 7.1s on a phone; the page weighs 2.4 MB, 1.5 MB of it images (about 966 KB saveable) plus a 280 KB animation script from unicorn.studio; real Chrome visitors pass Core Web Vitals (p75 LCP 0.72s, INP 95 ms, CLS 0); accessibility 86 with low-contrast text, out-of-order headings and links without names.Source
    • Your websiteHomepage (checked twice on November 10, 2026): title "Nucleo | Automated Cancer Diagnostics"; no H1 heading; between "Faster than manual segmentation." and "Agreement between Nucleo and experts." the page text reads "0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9%", and "2,500" and "98%" appear nowhere; the sections "We are backed by", "We work with the best hospitals in the US and worldwide" and "Our technology is powered by industry leaders" name no one in the text; the only call to action is a Cal.com "Book a demo" link.Source

This quarter 11

The core plays for the next three months.

  • 09

    Nucleo's registered study sets public thresholds (Dice 0.90 for muscle, subcutaneous and visceral fat; kappa 0.80 for sarcopenia), but its site has no validation page and the sitemap holds only the homepage. A competitor, iAID's Aid-U, has a published multicentre study reporting 97.4% agreement on 537 patients, which is what a buyer will compare against.

    High impactMedium effort First result in 1 month Founder & CTO

    First action

    Draft the results table from the study outputs.

    How to measure it

    Preprint downloads

    Now
    –
    Target
    500
    By
    90 days

    Steps

    1. 1

      Write the results

      Per tissue, against pre-set thresholds.

      Founder & CTO 1 week

    2. 2

      Clinical co-author

      A partner radiologist reviews and co-signs.

      Founder & CEO 2 weeks

    3. 3

      Post the preprint

      medRxiv or arXiv.

      Founder & CTO 1 day

    4. 4

      Build the page

      /validation with plots and method.

      Freelance web designer 3 days

    5. 5

      Link claims

      Homepage numbers point here.

      Freelance web designer 1 hour

    6. 6

      Submit

      To a radiology journal or conference.

      Founder & CTO 1 month

    First result
    1 month
    Cost
    Founder time; journal fees if open access
    Owner
    Founder & CTO

    Smallest test

    Publish the results table on LinkedIn.

    Why it works

    Publish a /validation page with the method, dataset, per-tissue Dice, Bland-Altman plots and kappa, plus a preprint. Add lesion sizing and RECIST results as they are ready. Link every claim on the homepage to it.

    Medical buyers and researchers cite papers, not landing pages. A preprint is also what AI assistants and search engines surface when someone asks which CT body-composition tools are validated.

    Risks and how to handle them

    • Public data differs from hospital data

      Say so, and plan a partner-site validation next.

    Builds on

    Evidence

    • ClinicalTrials.govNCT07600866, sponsored by Nucleo Research, Inc.: "Validation of a Body-Composition Segmentation Software (Soma)" on 200 publicly available CT scans; primary outcome Dice vs a multi-rater radiologist reference (thresholds 0.90 for muscle, subcutaneous and visceral fat, 0.85 intramuscular fat); secondary outcomes include Cohen's kappa for sarcopenia by SMI (threshold 0.80); status still "Not yet recruiting" with estimated start May 31, 2026 and primary completion June 15, 2026, last posted May 22, 2026.Source
    • Competitor researchA 2025 JMIR Formative Research study of iAID's Aid-U automated CT body-composition system: 537 patients at three institutions, 97.4% segmentation agreement with experts, images pulled from PACS automatically.Source
    • Your websiteSitemap: one URL, the homepage. There are no pages for individual use cases, validation results, publications, pricing, security or careers, and no blog.Source
  • 10

    Nucleo's whole site is one URL. Its three use cases each have a different buyer and search vocabulary ("CT body composition", "sarcopenia L3 SMI", "RECIST target lesion"), but they share a few lines on one page, so there is nothing for a researcher's search to land on.

    Medium impactMedium effort First result in 2 months Founder & CEO

    First action

    Write the body-composition page first, since it has the registered validation.

    How to measure it

    Use-case page visits

    Now
    –
    Target
    300 a month
    By
    2 quarters

    Steps

    1. 1

      Pick search terms

      How clinicians name each task.

      Founder & CEO 1 day

    2. 2

      Write page 1

      Body composition and sarcopenia.

      Founder & CEO 3 days

    3. 3

      Write pages 2 and 3

      Lesion sizing; RECIST classification.

      Founder & CTO 1 week

    4. 4

      Add outputs

      Sample images from idea 6.

      Freelance web designer 2 days

    5. 5

      Publish and submit

      Sitemap and Search Console.

      Freelance web designer 1 day

    6. 6

      Review

      Queries and demos per page.

      Founder & CEO monthly

    First result
    2 months
    Cost
    Designer time
    Owner
    Founder & CEO

    Smallest test

    Publish the body-composition page alone.

    Why it works

    Build three pages, each with the clinical problem, what Nucleo measures, a sample output, validation numbers, who uses it (radiology, surgical oncology, trials) and a demo link. Add each to the sitemap.

    Specific pages match specific searches and give partners something precise to forward to colleagues.

    Risks and how to handle them

    • Thin pages

      Include real outputs and numbers on each.

    Builds on

    Evidence

    • Your websiteSitemap: one URL, the homepage. There are no pages for individual use cases, validation results, publications, pricing, security or careers, and no blog.Source
    • Your websiteHomepage content: three use cases, "Body composition and sarcopenia assessment", "Tumor Lesion Sizing" and "Target vs Non-Target Lesion Classification" (RECIST); meta description "Advanced AI for oncological medical imaging. From scan to answer. In seconds, not hours."; no product name, no intended-use or regulatory statement, and no privacy, security or careers link; the footer links to Y Combinator, LinkedIn and X.Source
  • 11

    Nucleo's job post says pharma teams already use it, and RECIST target vs non-target classification is one of its three use cases. Trial imaging is a market where Mint Medical sells RECIST "automated compliance checks" to pharma and CROs and Median Technologies runs an imaging CRO for phases I-III. Trial analysis is a research setting, so it does not have to wait for clinical clearance in the way hospital diagnosis does.

    High impactMedium effort First result in 1 quarter Founder & CEO

    First action

    List ten imaging CROs and the pharma teams already using Nucleo.

    How to measure it

    CRO or pharma pilots

    Now
    –
    Target
    2
    By
    2 quarters

    Steps

    1. 1

      Define the offer

      Inputs, outputs, reader workflow, audit trail.

      Founder & CTO 1 week

    2. 2

      Price it

      Per timepoint, with a pilot rate.

      Founder & CEO 2 days

    3. 3

      Build a list

      Ten CROs and current pharma users.

      Founder & CEO 2 days

    4. 4

      Retro pilot

      One completed study, reader time and agreement.

      AI Engineer (open role) 1 month

    5. 5

      Write it up

      A one-page result for the next CRO.

      Founder & CEO 1 week

    6. 6

      Page

      A /clinical-trials page.

      Freelance web designer 3 days

    First result
    1 quarter
    Cost
    Founder and engineer time
    Owner
    Founder & CEO

    Smallest test

    Offer one CRO a free retrospective comparison on 50 timepoints.

    Why it works

    Package a trial offer: pre-read lesion sizing and target/non-target suggestions that a CRO's readers confirm, priced per timepoint, with an audit trail. Pilot with one CRO on a completed study, measuring reader time saved and agreement.

    CROs are paid on throughput and consistency, and the 2,500x segmentation speed matters most where thousands of timepoints are read.

    Risks and how to handle them

    • Validation requirements of trial sponsors

      Start on retrospective data and agree acceptance criteria up front.

    Builds on

    Evidence

    • HiringAI Engineer (San Francisco or remote, $70K-$120K, listed on Y Combinator): Nucleo turns "routine CT and MRI imaging into quantitative insights", works with "top-tier hospitals in the US and Europe" plus researchers and pharma teams, and is "building toward FDA-cleared clinical software"; the role enforces regulatory policies and tracks sign-offs, travels for on-site integrations and onboards new users; on-prem integration experience is a plus.Source
    • Competitor researchMint Medical: mint Lesion for structured oncology reads with RECIST/iRECIST "automated compliance checks" for clinical trials, pharma and CROs; names hospital users such as The Royal Marsden, Stanford University and German university hospitals.Source
    • Competitor researchMedian Technologies: an imaging CRO for oncology trials phase I-III plus eyonis LCS, which has FDA 510(k) clearance and CE marking; distributes through partners Tempus (U.S.) and Olea Medical.Source
  • 12

    Nucleo serves researchers as well as hospitals, and body composition from CT is a common research measure: published work shows CT finds sarcopenia far more often than BMI-based screening. Academic groups usually contour L3 slices by hand for months. Nucleo has no pricing page and sells only through demos.

    Medium impactMedium effort First result in 2 months Founder & CEO

    First action

    Set a price per scan and a minimum per study.

    How to measure it

    Research studies purchased

    Now
    –
    Target
    5
    By
    2 quarters

    Steps

    1. 1

      Set the price

      Per scan with a study minimum.

      Founder & CEO 1 day

    2. 2

      Write the agreement

      Data use and de-identification terms.

      Legal counsel 2 weeks

    3. 3

      Template the output

      Spreadsheet plus methods text.

      Founder & CTO 1 week

    4. 4

      Publish the page

      Price band and request form.

      Freelance web designer 2 days

    5. 5

      Tell partner labs

      Researchers at current partner hospitals.

      Founder & CEO 2 weeks

    6. 6

      Track citations

      Papers using Nucleo.

      AI Engineer (open role) quarterly

    First result
    2 months
    Cost
    Legal review
    Owner
    Founder & CEO

    Smallest test

    Offer one partner lab a quoted price for its next cohort.

    Why it works

    Offer a fixed-price research licence: upload a de-identified cohort under a data agreement (or run on-premises), get muscle, fat and SMI per patient in a spreadsheet, with a methods paragraph and citation. Publish the price band and a request form.

    Researchers have small budgets but quick decisions; a fixed price per study turns a long hospital sale into a purchase order, and every paper becomes a citation.

    Risks and how to handle them

    • Data transfer worries

      Offer an on-premises run, which the team already supports.

    Evidence

    • HiringAI Engineer (San Francisco or remote, $70K-$120K, listed on Y Combinator): Nucleo turns "routine CT and MRI imaging into quantitative insights", works with "top-tier hospitals in the US and Europe" plus researchers and pharma teams, and is "building toward FDA-cleared clinical software"; the role enforces regulatory policies and tracks sign-offs, travels for on-site integrations and onboards new users; on-prem integration experience is a plus.Source
    • Your websiteSitemap: one URL, the homepage. There are no pages for individual use cases, validation results, publications, pricing, security or careers, and no blog.Source
    • Competitor researchA 2025 JMIR Formative Research study of iAID's Aid-U automated CT body-composition system: 537 patients at three institutions, 97.4% segmentation agreement with experts, images pulled from PACS automatically.Source
  • 13

    Nucleo Research, Inc. registered on SAM.gov on October 15, 2026 for "All awards", under NAICS 541715 (research and development), and is listed as a woman-owned business. Registration is the entry ticket for federal grants and contracts, and the company's mission, shorter time to cancer treatment, fits federal cancer research programmes.

    Medium impactMedium effort First result in 6 months Founder & CEO

    First action

    List open solicitations that match the three use cases.

    How to measure it

    Federal proposals submitted

    Now
    –
    Target
    1
    By
    2 quarters

    Steps

    1. 1

      Check eligibility

      Programme rules and deadlines.

      Founder & CEO 2 days

    2. 2

      Pick a topic

      Match to a use case.

      Founder & CTO 1 week

    3. 3

      Line up a site

      A partner hospital as clinical collaborator.

      Founder & CEO 2 weeks

    4. 4

      Write

      Aims, validation and commercial plan.

      Founder & CEO 1 month

    5. 5

      Review

      An experienced grant reviewer reads it.

      Grant consultant 1 week

    6. 6

      Submit

      On the programme deadline.

      Founder & CEO 1 day

    First result
    6 months
    Cost
    Founder time; optional grant consultant
    Owner
    Founder & CEO

    Smallest test

    Read one open solicitation this week and score the fit.

    Why it works

    Check eligibility for small-business research programmes (for example SBIR/STTR at the National Cancer Institute), pick the one whose topic fits body composition or treatment response, and write a proposal with a partner hospital as the clinical site.

    Non-dilutive funding pays for the validation and regulatory work that hospitals and FDA will ask for, and a federal award is itself a credibility signal.

    Risks and how to handle them

    • Time away from selling

      Reuse the validation write-up (idea 9) as the core.

    Builds on

    Evidence

    • Public recordsSAM.gov: NUCLEO RESEARCH, INC. registered October 15, 2026 (active to October 15, 2027) for "All awards"; NAICS 541715 (research and development) and 541511; listed as a woman-owned business; Delaware corporation, San Francisco.Source
    • Launch YCLaunch YC post: calls cancer "a logistical failure"; median U.S. diagnostic interval about 36 days and time to treatment about 38 days, so 9-11 weeks before therapy; cites a meta-analysis estimating a 6-13% rise in relative mortality per four-week treatment delay; a doctor imports a CT scan and Nucleo extracts sarcopenia, body composition, lesion volume and target vs non-target classification.Source
  • 14

    Nucleo's launch post makes a strong case: a median diagnostic interval of about 36 days plus about 38 days to treatment, 9-11 weeks in total, and a 6-13% rise in relative mortality per four weeks of delay. None of this is on the website, which talks about speed in seconds rather than weeks saved for patients.

    Medium impactLow effort First result in 1 month Founder & CEO

    First action

    Collect the cited studies from the launch post into one source list.

    How to measure it

    Demos from the page

    Now
    –
    Target
    5 a month
    By
    90 days

    Steps

    1. 1

      Gather sources

      Each statistic with its citation.

      Founder & CEO 1 day

    2. 2

      Map the pathway

      Scan to treatment, with delays.

      Founder & CEO 2 days

    3. 3

      Clinical review

      A partner oncologist checks it.

      Founder & CEO 1 week

    4. 4

      Build the page

      Diagram, sources, estimate.

      Freelance web designer 3 days

    5. 5

      Share

      LinkedIn, X and outreach emails.

      Founder & CEO 1 day

    6. 6

      Measure

      Shares and demos.

      Founder & CEO monthly

    First result
    1 month
    Cost
    Designer time
    Owner
    Founder & CEO

    Smallest test

    Post the pathway diagram on LinkedIn.

    Why it works

    Publish a sourced page, "Where the weeks go", that maps the steps from scan to first treatment, shows which ones imaging measurement slows down, and where Nucleo helps. Add a simple estimate a cancer centre can fill in for its own volume.

    Administrators and oncology leaders buy outcomes such as faster treatment starts; a clear, cited page gives a clinical champion something to forward to budget holders.

    Risks and how to handle them

    • Over-claiming Nucleo's effect on mortality

      Show the problem with sources; describe only measured time saved.

    Evidence

    • Launch YCLaunch YC post: calls cancer "a logistical failure"; median U.S. diagnostic interval about 36 days and time to treatment about 38 days, so 9-11 weeks before therapy; cites a meta-analysis estimating a 6-13% rise in relative mortality per four-week treatment delay; a doctor imports a CT scan and Nucleo extracts sarcopenia, body composition, lesion volume and target vs non-target classification.Source
    • Your websiteSitemap: one URL, the homepage. There are no pages for individual use cases, validation results, publications, pricing, security or careers, and no blog.Source
  • 15

    Hospital deals pass through IT security and privacy review. Nucleo's site has no security, privacy or deployment page, while its job post shows it already does on-site and on-premises integrations. The answers exist but are given one questionnaire at a time.

    Medium impactMedium effort First result in 2 months Founder & CTO

    First action

    Collect answers from the last two hospital security questionnaires.

    How to measure it

    Time to go-live

    Now
    –
    Target
    -30%
    By
    2 quarters

    Steps

    1. 1

      Collect answers

      Past questionnaires.

      Founder & CTO 2 days

    2. 2

      Draft the page

      Deployment, data flow, controls.

      Founder & CTO 3 days

    3. 3

      Privacy policy

      Publish and link in the footer.

      Legal counsel 1 week

    4. 4

      Security summary

      Downloadable PDF.

      Founder & CTO 2 days

    5. 5

      Publish

      /security, linked from demos.

      Freelance web designer 1 day

    6. 6

      Keep updated

      After each new deployment type.

      AI Engineer (open role) quarterly

    First result
    2 months
    Cost
    Legal review
    Owner
    Founder & CTO

    Smallest test

    Send the draft summary with the next pilot proposal.

    Why it works

    Publish a page covering deployment options (on-premises, private cloud), where images go, de-identification, access control, logging, and how HIPAA obligations are handled, with a downloadable security summary and a contact for questionnaires.

    Answering the first ten security questions in public shortens the slowest stage of a hospital sale and shows maturity beyond team size.

    Risks and how to handle them

    • Promising controls not yet in place

      Only describe what exists; list the roadmap separately.

    Evidence

    • Your websiteHomepage content: three use cases, "Body composition and sarcopenia assessment", "Tumor Lesion Sizing" and "Target vs Non-Target Lesion Classification" (RECIST); meta description "Advanced AI for oncological medical imaging. From scan to answer. In seconds, not hours."; no product name, no intended-use or regulatory statement, and no privacy, security or careers link; the footer links to Y Combinator, LinkedIn and X.Source
    • HiringAI Engineer (San Francisco or remote, $70K-$120K, listed on Y Combinator): Nucleo turns "routine CT and MRI imaging into quantitative insights", works with "top-tier hospitals in the US and Europe" plus researchers and pharma teams, and is "building toward FDA-cleared clinical software"; the role enforces regulatory policies and tracks sign-offs, travels for on-site integrations and onboards new users; on-prem integration experience is a plus.Source
  • 16

    The launch post describes the workflow as "a doctor imports a CT scan". A competitor's published study highlights that its body-composition system pulls images from PACS automatically, and Nucleo's engineers already travel for on-site integrations. Buyers cannot see on the site how results reach the radiologist or the oncologist.

    Medium impactMedium effort First result in 1 quarter Founder & CTO

    First action

    Write down how each current site gets scans in and results out.

    How to measure it

    Weekly scans per site

    Now
    –
    Target
    x3
    By
    2 quarters

    Steps

    1. 1

      Map current sites

      Import and export paths.

      Founder & CTO 2 days

    2. 2

      Pick a standard path

      PACS pull, structured result back.

      Founder & CTO 1 week

    3. 3

      Build

      Connector and report output.

      AI Engineer (open role) 1 month

    4. 4

      Pilot

      At one partner site.

      AI Engineer (open role) 1 month

    5. 5

      Document

      Diagram and integration guide.

      Founder & CTO 1 week

    6. 6

      Publish

      /how-it-works page.

      Freelance web designer 2 days

    First result
    1 quarter
    Cost
    Engineering time
    Owner
    Founder & CTO

    Smallest test

    Publish the current workflow diagram.

    Why it works

    Document and, where possible, productise the integration: automatic retrieval from PACS, results returned as a structured report or DICOM object, and an export for research. Publish a one-page diagram and a short integration guide.

    Manual import is the biggest drag on daily use; automatic flow raises scans per site and makes the per-site usage numbers (idea 17) grow on their own.

    Risks and how to handle them

    • Every hospital differs

      Standardise one path and treat others as projects.

    Builds on

    Evidence

    • Launch YCLaunch YC post: calls cancer "a logistical failure"; median U.S. diagnostic interval about 36 days and time to treatment about 38 days, so 9-11 weeks before therapy; cites a meta-analysis estimating a 6-13% rise in relative mortality per four-week treatment delay; a doctor imports a CT scan and Nucleo extracts sarcopenia, body composition, lesion volume and target vs non-target classification.Source
    • Competitor researchA 2025 JMIR Formative Research study of iAID's Aid-U automated CT body-composition system: 537 patients at three institutions, 97.4% segmentation agreement with experts, images pulled from PACS automatically.Source
    • HiringAI Engineer (San Francisco or remote, $70K-$120K, listed on Y Combinator): Nucleo turns "routine CT and MRI imaging into quantitative insights", works with "top-tier hospitals in the US and Europe" plus researchers and pharma teams, and is "building toward FDA-cleared clinical software"; the role enforces regulatory policies and tracks sign-offs, travels for on-site integrations and onboards new users; on-prem integration experience is a plus.Source
  • 17

    Nucleo's main claim is speed (2,500x faster than manual segmentation), and its job post includes onboarding and supporting users. But hospital partners only see the value if someone counts it. A pilot without a number at renewal time is easy to drop.

    Medium impactLow effort First result in 1 month Founder & CEO

    First action

    Agree a manual-time baseline per task with one partner.

    How to measure it

    Pilots converted to paid

    Now
    –
    Target
    50%
    By
    2 quarters

    Steps

    1. 1

      Set baselines

      Manual minutes per task, agreed with clinicians.

      Founder & CEO 1 week

    2. 2

      Log usage

      Scans and tasks per site.

      AI Engineer (open role) 1 week

    3. 3

      Template

      One-page report.

      Founder & CEO 1 day

    4. 4

      Send monthly

      To the champion and their lead.

      Founder & CEO monthly

    5. 5

      Quarterly review

      Value and next department.

      Founder & CEO quarterly

    6. 6

      Aggregate

      Total hours saved for the homepage.

      Founder & CEO quarterly

    First result
    1 month
    Cost
    $0
    Owner
    Founder & CEO

    Smallest test

    Send one partner a report this month.

    Why it works

    Each month, send each site a one-page report: scans processed, tasks run, estimated manual time avoided (using an agreed manual-time baseline), turnaround times and any discrepancies flagged. Use it in quarterly reviews and expansion talks.

    Visible value turns pilots into contracts and gives the internal champion a number to show their department.

    Risks and how to handle them

    • Estimates questioned

      Show the baseline and method on the report.

    Evidence

    • Y Combinator profileFall 2025 batch, team of 2, San Francisco; "The first Agentic platform for Oncology"; segmentation "2,500x faster" than manual, "state-of-the-art performance in sarcopenia detection", "98% agreement between Nucleo and medical experts" in validation studies; aims to cut the 9-11 week diagnosis-to-treatment window to 10 days or less; hospital partners listed: Stanford Hospital, Cedars-Sinai, UCI Health and Weill Cornell.Source
    • HiringAI Engineer (San Francisco or remote, $70K-$120K, listed on Y Combinator): Nucleo turns "routine CT and MRI imaging into quantitative insights", works with "top-tier hospitals in the US and Europe" plus researchers and pharma teams, and is "building toward FDA-cleared clinical software"; the role enforces regulatory policies and tracks sign-offs, travels for on-site integrations and onboards new users; on-prem integration experience is a plus.Source
  • 18

    Sarcopenia and body composition are one of Nucleo's three use cases, and the company claims state-of-the-art sarcopenia detection. In practice those numbers matter most to the teams planning surgery, chemotherapy dosing and nutrition support, not only to radiology. Each partner hospital is several potential buyers.

    Medium impactLow effort First result in 2 months Founder & CEO

    First action

    Ask one champion for one introduction this month.

    How to measure it

    Departments per hospital

    Now
    –
    Target
    2
    By
    2 quarters

    Steps

    1. 1

      Map departments

      Who uses sarcopenia data at each partner.

      Founder & CEO 1 week

    2. 2

      Ask champions

      Two introductions each.

      Founder & CEO 2 weeks

    3. 3

      Tailor the demo

      Surgery and nutrition use cases.

      Founder & CTO 1 week

    4. 4

      Run demos

      Short, on their pathway.

      Founder & CEO 1 month

    5. 5

      Propose joint projects

      Radiology plus the new team.

      Founder & CEO ongoing

    6. 6

      Thank champions

      Co-authorship or speaking slots.

      Founder & CEO ongoing

    First result
    2 months
    Cost
    $0
    Owner
    Founder & CEO

    Smallest test

    One introduction at one partner.

    Why it works

    In every partner hospital, ask the radiology champion for two introductions: surgical oncology and clinical nutrition (or geriatric oncology). Offer each a short demo using sarcopenia outputs on their own patient pathway, and a joint project with radiology.

    A warm internal introduction is the cheapest sale a small company can make, and it spreads usage across departments, which makes the contract stickier.

    Risks and how to handle them

    • Champions protective of their project

      Frame it as shared credit and a joint study.

    Evidence

    • Your websiteHomepage content: three use cases, "Body composition and sarcopenia assessment", "Tumor Lesion Sizing" and "Target vs Non-Target Lesion Classification" (RECIST); meta description "Advanced AI for oncological medical imaging. From scan to answer. In seconds, not hours."; no product name, no intended-use or regulatory statement, and no privacy, security or careers link; the footer links to Y Combinator, LinkedIn and X.Source
    • Y Combinator profileFall 2025 batch, team of 2, San Francisco; "The first Agentic platform for Oncology"; segmentation "2,500x faster" than manual, "state-of-the-art performance in sarcopenia detection", "98% agreement between Nucleo and medical experts" in validation studies; aims to cut the 9-11 week diagnosis-to-treatment window to 10 days or less; hospital partners listed: Stanford Hospital, Cedars-Sinai, UCI Health and Weill Cornell.Source
  • 19

    Nucleo works with researchers and has a registered validation behind its body-composition software. A competitor's strongest marketing asset is a published multicentre study. Every paper that uses Nucleo's measurements and names it in the methods is a citation that other researchers find and copy.

    Medium impactMedium effort First result in 2 quarters Founder & CTO

    First action

    Invite researchers at current partner hospitals to propose one study each.

    How to measure it

    Publications naming Nucleo

    Now
    –
    Target
    3 submitted
    By
    3 quarters

    Steps

    1. 1

      Set terms

      Scope, data rules, credit.

      Founder & CEO 1 week

    2. 2

      Invite proposals

      Partner hospitals first.

      Founder & CEO 2 weeks

    3. 3

      Select

      Two or three strongest.

      Founder & CTO 1 week

    4. 4

      Run analyses

      Under data agreements.

      AI Engineer (open role) 1 month

    5. 5

      Methods text

      Provide a standard description.

      Founder & CTO 1 day

    6. 6

      Publications page

      List each paper.

      Freelance web designer 1 day

    First result
    2 quarters
    Cost
    Engineering time
    Owner
    Founder & CTO

    Smallest test

    One study with one partner lab.

    Why it works

    Offer a limited number of research groups free or discounted analysis for a defined study, in exchange for naming Nucleo in the methods and a co-author or acknowledgement. Track each paper and list them on a /publications page.

    Methods sections are how research tools spread; a list of papers is durable proof for hospitals, pharma and grant reviewers.

    Risks and how to handle them

    • Free work crowds out paid

      Cap at three studies a year; use the paid research licence (idea 12) for the rest.

    Builds on

    Evidence

    • HiringAI Engineer (San Francisco or remote, $70K-$120K, listed on Y Combinator): Nucleo turns "routine CT and MRI imaging into quantitative insights", works with "top-tier hospitals in the US and Europe" plus researchers and pharma teams, and is "building toward FDA-cleared clinical software"; the role enforces regulatory policies and tracks sign-offs, travels for on-site integrations and onboards new users; on-prem integration experience is a plus.Source
    • Competitor researchA 2025 JMIR Formative Research study of iAID's Aid-U automated CT body-composition system: 537 patients at three institutions, 97.4% segmentation agreement with experts, images pulled from PACS automatically.Source
    • ClinicalTrials.govNCT07600866, sponsored by Nucleo Research, Inc.: "Validation of a Body-Composition Segmentation Software (Soma)" on 200 publicly available CT scans; primary outcome Dice vs a multi-rater radiologist reference (thresholds 0.90 for muscle, subcutaneous and visceral fat, 0.85 intramuscular fat); secondary outcomes include Cohen's kappa for sarcopenia by SMI (threshold 0.80); status still "Not yet recruiting" with estimated start May 31, 2026 and primary completion June 15, 2026, last posted May 22, 2026.Source

Big bets 3

Larger plays that can change the trajectory.

  • 20

    Nucleo says it is "building toward FDA-cleared clinical software" and is hiring an engineer to run regulatory sign-offs. It has three use cases, but only body composition has a registered, pre-specified validation. Median Technologies shows what clearance buys: its FDA-cleared product anchors its commercial partnerships.

    High impactHigh effort First result in 12+ months Founder & CTO

    First action

    Hire or contract a regulatory consultant to assess the pathway for body composition.

    How to measure it

    First FDA submission

    Now
    –
    Target
    submitted
    By
    12 months

    Steps

    1. 1

      Pathway assessment

      Device classification and route for body composition.

      Regulatory consultant 1 month

    2. 2

      Quality system

      Design controls and documentation.

      AI Engineer (open role) 3 months

    3. 3

      Pre-submission

      Request FDA feedback on the plan.

      Regulatory consultant 3 months

    4. 4

      Clinical validation

      Multi-site data with partner hospitals.

      Founder & CTO 6 months

    5. 5

      Submit

      The first application.

      Regulatory consultant 1 month

    6. 6

      Prepare launch

      Clinical labelling, pricing and site plan.

      Founder & CEO 3 months

    First result
    12+ months
    Cost
    Regulatory consulting and validation costs; a large share of a seed budget
    Owner
    Founder & CTO

    Smallest test

    A one-month pathway assessment.

    Why it works

    Pick the narrowest first indication with the strongest evidence (body-composition measurement from CT), agree the regulatory pathway with an experienced consultant, request FDA feedback through a pre-submission meeting, and build the quality system around that one product. Add lesion sizing and RECIST later.

    A first clearance changes the sales conversation at every hospital from "research" to "clinical", and a narrow scope is faster and cheaper to get through.

    Risks and how to handle them

    • Cost and time overrun

      Fund through research revenue (ideas 11, 12) and federal grants (idea 13).

    Builds on

    Evidence

    • HiringAI Engineer (San Francisco or remote, $70K-$120K, listed on Y Combinator): Nucleo turns "routine CT and MRI imaging into quantitative insights", works with "top-tier hospitals in the US and Europe" plus researchers and pharma teams, and is "building toward FDA-cleared clinical software"; the role enforces regulatory policies and tracks sign-offs, travels for on-site integrations and onboards new users; on-prem integration experience is a plus.Source
    • ClinicalTrials.govNCT07600866, sponsored by Nucleo Research, Inc.: "Validation of a Body-Composition Segmentation Software (Soma)" on 200 publicly available CT scans; primary outcome Dice vs a multi-rater radiologist reference (thresholds 0.90 for muscle, subcutaneous and visceral fat, 0.85 intramuscular fat); secondary outcomes include Cohen's kappa for sarcopenia by SMI (threshold 0.80); status still "Not yet recruiting" with estimated start May 31, 2026 and primary completion June 15, 2026, last posted May 22, 2026.Source
    • Competitor researchMedian Technologies: an imaging CRO for oncology trials phase I-III plus eyonis LCS, which has FDA 510(k) clearance and CE marking; distributes through partners Tempus (U.S.) and Olea Medical.Source
  • 21

    With a team of two and a single "Book a demo" link, Nucleo sells one hospital at a time. Median Technologies, much larger, still distributes its cleared product through partners (Tempus in the U.S., Olea Medical). Imaging AI is often bought through platforms that hospitals already use.

    High impactHigh effort First result in 6 months Founder & CEO

    First action

    List ten platforms where current partner hospitals already buy imaging AI.

    How to measure it

    Partner-sourced sites

    Now
    –
    Target
    10
    By
    12 months

    Steps

    1. 1

      Shortlist

      Platforms partners already use.

      Founder & CEO 2 weeks

    2. 2

      Qualify

      Integration terms and revenue share.

      Founder & CEO 1 month

    3. 3

      Pick one

      Best fit and speed.

      Founder & CEO 1 week

    4. 4

      Integrate

      Technical connection and listing.

      Founder & CTO 3 months

    5. 5

      Joint launch

      Webinar and case study.

      Founder & CEO 1 month

    6. 6

      Review

      Sites and revenue per quarter.

      Founder & CEO quarterly

    First result
    6 months
    Cost
    Engineering time and revenue share
    Owner
    Founder & CEO

    Smallest test

    One exploratory call with one platform.

    Why it works

    Shortlist imaging AI platforms, PACS vendors and oncology data companies whose customers need body composition and lesion measurement. Propose a pilot integration with one, starting with research customers, then clinical once cleared.

    One integration can expose Nucleo to hundreds of sites that already passed procurement for the partner, which a two-person team cannot reach directly.

    Risks and how to handle them

    • Partner controls the customer

      Keep direct relationships with clinical champions.

    Builds on

    Evidence

    • Competitor researchMedian Technologies: an imaging CRO for oncology trials phase I-III plus eyonis LCS, which has FDA 510(k) clearance and CE marking; distributes through partners Tempus (U.S.) and Olea Medical.Source
    • Y Combinator profileFall 2025 batch, team of 2, San Francisco; "The first Agentic platform for Oncology"; segmentation "2,500x faster" than manual, "state-of-the-art performance in sarcopenia detection", "98% agreement between Nucleo and medical experts" in validation studies; aims to cut the 9-11 week diagnosis-to-treatment window to 10 days or less; hospital partners listed: Stanford Hospital, Cedars-Sinai, UCI Health and Weill Cornell.Source
    • Your websiteHomepage (checked twice on November 10, 2026): title "Nucleo | Automated Cancer Diagnostics"; no H1 heading; between "Faster than manual segmentation." and "Agreement between Nucleo and experts." the page text reads "0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9%", and "2,500" and "98%" appear nowhere; the sections "We are backed by", "We work with the best hospitals in the US and worldwide" and "Our technology is powered by industry leaders" name no one in the text; the only call to action is a Cal.com "Book a demo" link.Source
  • 22

    Pharma teams already use Nucleo, and its body-composition software has registered thresholds for muscle and fat segmentation. Muscle loss during cancer treatment matters for drug tolerance and outcomes, so sponsors want it measured consistently across sites, from CT scans trials already collect. The trial-imaging competitors (Mint Medical, Median Technologies) focus on tumour response, not body composition.

    High impactHigh effort First result in 9 months Founder & CEO

    First action

    Ask current pharma users whether body composition is measured in their trials today, and how.

    How to measure it

    Pharma trial contracts

    Now
    –
    Target
    1
    By
    12 months

    Steps

    1. 1

      Discovery calls

      Five pharma or biotech teams.

      Founder & CEO 1 month

    2. 2

      Define the service

      Measures, QC, report format.

      Founder & CTO 1 month

    3. 3

      Validation dossier

      Method and reproducibility package.

      Founder & CTO 2 months

    4. 4

      Retrospective pilot

      One completed trial.

      AI Engineer (open role) 3 months

    5. 5

      Price

      Per timepoint and setup.

      Founder & CEO 1 week

    6. 6

      Partner with CROs

      Offer through trial imaging providers.

      Founder & CEO ongoing

    First result
    9 months
    Cost
    Engineering and validation time
    Owner
    Founder & CEO

    Smallest test

    One retrospective analysis for one sponsor.

    Why it works

    Build a trial service: standardised muscle and fat measures at every CT timepoint, central quality control, a statistical report and a validation dossier sponsors can file. Start with retrospective analysis of a completed trial for one sponsor.

    Trial endpoints are high-value, recurring and multi-year, and data from many sites builds the evidence Nucleo needs for clinical claims.

    Risks and how to handle them

    • Long sponsor sales cycles

      Start with retrospective analyses that need no protocol change.

    Builds on

    Evidence

    • HiringAI Engineer (San Francisco or remote, $70K-$120K, listed on Y Combinator): Nucleo turns "routine CT and MRI imaging into quantitative insights", works with "top-tier hospitals in the US and Europe" plus researchers and pharma teams, and is "building toward FDA-cleared clinical software"; the role enforces regulatory policies and tracks sign-offs, travels for on-site integrations and onboards new users; on-prem integration experience is a plus.Source
    • ClinicalTrials.govNCT07600866, sponsored by Nucleo Research, Inc.: "Validation of a Body-Composition Segmentation Software (Soma)" on 200 publicly available CT scans; primary outcome Dice vs a multi-rater radiologist reference (thresholds 0.90 for muscle, subcutaneous and visceral fat, 0.85 intramuscular fat); secondary outcomes include Cohen's kappa for sarcopenia by SMI (threshold 0.80); status still "Not yet recruiting" with estimated start May 31, 2026 and primary completion June 15, 2026, last posted May 22, 2026.Source
    • Competitor researchMint Medical: mint Lesion for structured oncology reads with RECIST/iRECIST "automated compliance checks" for clinical trials, pharma and CROs; names hospital users such as The Royal Marsden, Stanford University and German university hospitals.Source

Wildcards 2

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

  • 23

    Nucleo's only channels are LinkedIn and X, with no newsletter, YouTube or blog, and the site has one page. Radiologists and oncology researchers respond to images; a segmentation on a real (public) scan shows the product in a way a claim cannot.

    Medium impactLow effort First result in 1 month Founder & CEO

    First action

    Prepare four posts from the validation cohort.

    How to measure it

    Demos from social

    Now
    –
    Target
    4 a month
    By
    2 quarters

    Steps

    1. 1

      Pick cases

      Public scans with clear teaching points.

      Founder & CTO 1 day

    2. 2

      Make the overlay

      Segmentation image plus values.

      Founder & CTO 1 hour a week

    3. 3

      Write the point

      One clinical line, reviewed.

      Founder & CEO 30 minutes a week

    4. 4

      Post

      LinkedIn and X, same day.

      Founder & CEO weekly

    5. 5

      Invite clinicians

      Ask partners to comment or guest-post.

      Founder & CEO monthly

    6. 6

      Review

      Engagement and demos.

      Founder & CEO monthly

    First result
    1 month
    Cost
    $0
    Owner
    Founder & CEO

    Smallest test

    Four weekly posts, then review.

    Why it works

    Each week, post one public CT case: Nucleo's segmentation overlay, the measured values, a one-line clinical point (for example how low muscle mass changes a treatment discussion) and a link to the sample report or validation page.

    A steady visual series builds a following among exactly the clinicians who buy, and each post is reusable on the website and in outreach.

    Risks and how to handle them

    • Implied clinical advice

      Use public, de-identified data and educational wording only.

    Builds on

    Evidence

    • Your websiteSite record: domain registered August 2024; the homepage headline "Nucleo | Automated Cancer Diagnostics" is unchanged since at least November 2025; the site links only LinkedIn and X, with no newsletter, YouTube channel or blog.Source
    • ClinicalTrials.govNCT07600866, sponsored by Nucleo Research, Inc.: "Validation of a Body-Composition Segmentation Software (Soma)" on 200 publicly available CT scans; primary outcome Dice vs a multi-rater radiologist reference (thresholds 0.90 for muscle, subcutaneous and visceral fat, 0.85 intramuscular fat); secondary outcomes include Cohen's kappa for sarcopenia by SMI (threshold 0.80); status still "Not yet recruiting" with estimated start May 31, 2026 and primary completion June 15, 2026, last posted May 22, 2026.Source
  • 24

    Nucleo's job post says it works with "top-tier hospitals in the US and Europe", but the website is US-only in tone and names no European partner. Mint Medical's homepage shows how much weight named European university hospitals carry with trial and research buyers.

    Medium impactMedium effort First result in 2 quarters Founder & CEO

    First action

    Ask the European partner whether a second hospital would join a shared protocol.

    How to measure it

    European network sites

    Now
    –
    Target
    3
    By
    3 quarters

    Steps

    1. 1

      Map partners

      Current European contacts.

      Founder & CEO 1 week

    2. 2

      Draft a protocol

      Shared retrospective body-composition study.

      Founder & CTO 1 month

    3. 3

      Recruit sites

      Two more hospitals.

      Founder & CEO 2 months

    4. 4

      Data agreements

      GDPR-compliant terms with each site.

      Legal counsel 2 months

    5. 5

      Joint abstract

      European radiology meeting.

      Founder & CTO 3 months

    6. 6

      European page

      Network and results.

      Freelance web designer 1 week

    First result
    2 quarters
    Cost
    Travel and legal costs
    Owner
    Founder & CEO

    Smallest test

    One shared protocol with one European partner.

    Why it works

    Turn the European relationships into a small research network: a shared body-composition protocol across two or three hospitals, a joint abstract at a European radiology meeting, and a European page. Use it as multi-country evidence for both FDA and later European regulatory work.

    Multi-country data makes validation stronger, and a research network builds reputation without waiting for clinical clearance.

    Risks and how to handle them

    • Spreading a two-person team thin

      Only start once the US pilots run without daily founder time.

    Builds on

    Evidence

    • HiringAI Engineer (San Francisco or remote, $70K-$120K, listed on Y Combinator): Nucleo turns "routine CT and MRI imaging into quantitative insights", works with "top-tier hospitals in the US and Europe" plus researchers and pharma teams, and is "building toward FDA-cleared clinical software"; the role enforces regulatory policies and tracks sign-offs, travels for on-site integrations and onboards new users; on-prem integration experience is a plus.Source
    • Competitor researchMint Medical: mint Lesion for structured oncology reads with RECIST/iRECIST "automated compliance checks" for clinical trials, pharma and CROs; names hospital users such as The Royal Marsden, Stanford University and German university hospitals.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.

ProgrammableDigital

Show it, don't just describe it

These ideas need something to show: a demo, a video, an animation, graphics or a deck. Digital turns screenshots or the site into all of them, in the company's own brand.

Make a demo (opens programmable.digital 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.

  • Strella growth strategyAn AI customer research platform that recruits participants, runs AI-moderated voice and video interviews, and synthesises themes, clips…22 growth ideas
  • Kids Dental Brands growth strategyA doctor-owned group of children's dental and orthodontic practices, founded in 2002 and based in Phoenix, with shared services (call…22 growth ideas
  • ZAGG growth strategyMobile accessories under the InvisibleShield, ZAGG and mophie brands: screen and camera protection, cases, tablet keyboards, chargers and…23 growth ideas
  • Ambient.ai growth strategyAgentic physical security software: Ambient Pulsar, a reasoning vision-language model, runs on existing cameras and access control systems…23 growth ideas
  • Sofar Ocean growth strategyAn ocean intelligence company: the Spotter sensor network (buoys, Scout USV, A-Size floats), marine weather forecasts built on its own…23 growth ideas
  • Cardless growth strategyAn embedded credit card platform: APIs, embedded components and a ready-made cardholder app that let brands launch their own credit cards…22 growth ideas
  • Soria growth strategyAn AI-native financial research terminal for sector-specialist investors, starting with healthcare: agents turn SEC, CMS…23 growth ideas
  • TeamDynamix growth strategyA unified no-code platform for IT service management with AI, plus IT asset management (including Sassafras), digital employee experience…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. Nucleo did not commission, review or endorse it, and Programmable is not affiliated with Nucleo. 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 Nucleo runs its business.

Something wrong, or do you represent Nucleo? 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 308 ideas

    Make the proof visible: numbers in text, named hospitals, one positioning line with intended use, the updated registry record, a sample report, lighter images and accessibility fixes.

  2. Days 31 to 609 ideas

    Build the evidence and buyer pages: Soma results, use-case pages, the time-to-treatment page, security and workflow pages, usage reports, referrals inside hospitals, research collaborations and weekly scan posts.

  3. Days 61 to 907 ideas

    Open revenue that does not wait for clearance (trial reads, research licences, federal research funding), then commit to the big bets and a European pilot.

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 segmentation that takes seconds instead of manual contouring. The most cited problem: weeks between diagnosis and the start of treatment.

  • Consistent tumour measurements and RECIST classification
  • Slow, variable manual measurement
  • In their words: “time to treatment”

Market

Competitive landscape

4 competitors compared: Mint Medical, Median Technologies, AI Metrics and iAID (Aid-U).

  • Gap to exploit: Body composition, lesion sizing and RECIST classification in one tool
  • Mint Medical: Reading workflow, not automated segmentation
  • Median Technologies: Lung and liver focus, not body composition

Hiring signals

Nucleo has one open role listed on Y Combinator, an AI Engineer, and it says a lot about the next year.

  • AI Engineer: Regulatory sign-offs toward FDA clearance, on-site integrations, hospital, research and pharma users. Ideas 3, 11, 15, 16, 20.

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