Summary
New to clinical trials? Sections i and ii explain how a trial runs and where the money is lost, in plain English, before any of the numbers below.
Live at Emory with a completed GxP audit. $340K in signed contracts and 37 pipeline trials in six months, on under $70K of total burn.
ICH E6(R3) is in force and the FDA launched Real-Time Clinical Trials in April 2026. The manual data layer has to change.
$20B+ a year for the product we sell today (~20,000 trials a year at roughly $1M of data work each), with adjacent products reaching far larger markets.
80-85% gross margin at steady state. Break-even at about 11 trials a year (~$1.25M annual fixed cost at full team).
Works with any EDC, sells through exclusive CRO partners, and builds a cross-sponsor dataset.
What a clinical trial is, and how one runs.
Written for readers who have not worked in clinical research. Every term used later in this brief is defined here, and again in the glossary in the appendix.
A controlled experiment that tests whether a drug or device is safe and whether it works. Nothing reaches patients without one. A large trial takes 3-6 years and costs tens of millions of dollars.
The company that owns the drug is the sponsor. It usually hires a CRO (contract research organization) to run the operational work, and the trial itself happens at sites: hospitals and clinics, 50-150 of them on a large study.
A trial's value is not the data. It is the date the data is trusted, because that is when the sponsor can file with the regulator and the product can start earning.
The six stages of a trial
- The protocol is written: the study's rulebook, saying who can enroll, what is measured, when, and in what units
- Separate documents give their own instructions: lab manual, pharmacy manual, consent form, questionnaires, output specification
- Each hospital is trained and activated
- Every site has its own record system, its own labs and its own way of working, so 60 sites means 60 different setups
- Sites recruit participants and take consent
- Usually the slowest stage, and the one sponsors most want to start earlier
- Blood tests, scans, vital signs, questionnaires and wearable readings are taken at each visit
- A coordinator then re-types those values by hand into the sponsor's trial database, the EDC (electronic data capture system, sold by Medidata, Veeva and Oracle)
- Monitors check the database for values that look wrong
- Each question raised is a query: a formal request back to the site to check or correct one value. A large study produces roughly 97,000 of them
- Remaining queries are cleared and the database is frozen: database lock
- The sponsor files with the FDA or the EMA and, if approved, can sell the product
- The measuring is advanced; moving the numbers between systems is done by people
- The gap between a measurement being taken and it being usable is where the money goes
- The EDC vendor sells a database and is not responsible for what happens before data arrives
- The CRO sells people, so more manual work is more revenue
- The site is measured on enrollment, not on data quality
Terms in bold are defined again in the glossary in the appendix.
The official path is tidy. Every arrow is a person.
Add the gaps up and you have the problem. The stages are fine. The time, the effort and the money sit in the handovers between them: about 8 days to get a value in, tens of thousands of questions that mostly change nothing, and 36 days on average from the last patient visit to a locked database, at $555.7K for every day it runs late: $55.7K of direct cost plus ~$500K of forgone sales. And the delay is not only financial: those are days a treatment that works is not yet available to the patients waiting for it.
The same handovers, with the sources
| Handover | What actually happens | What it costs |
|---|---|---|
| Site to database | A person reads a value off a hospital screen or a lab report and types it into a different system | 8.1 days to arrive 3 8.23% typed wrong 5 |
| Database to review | Monitors look for values that contradict the protocol | ~97,000 queries 58% change nothing 6 |
| Review back to site | Each query is emailed, answered, checked and closed by hand | 9.6 days each 16.4 if held elsewhere |
| Last visit to lock | The backlog of unresolved problems is worked through | 36 days on average 3 |
| Every day late | The trial keeps running and the product is not yet earning | $555.7K per day 7 $55.7K direct + ~$500K forgone sales |
And a quieter problem upstream. The protocol, the lab manual, the consent form and the output specification are written and approved separately, and nobody reconciles them against each other. The protocol says kilograms, the lab reports pounds. The protocol allows a visit window of three days, the schedule says seven. These conflicts cost nothing on the day they are written and surface in month eight as queries, deviations and data that cannot be used.
This is the layer Nightingale replaces: not the science, not the database, but the manual handling between the source and the sponsor.
Trial data is still re-keyed and cleaned by hand.
Regulators now expect traceable, real-time trial data.
In April 2026 the FDA launched Real-Time Clinical Trials, with data sent to regulators continuously instead of at database lock.
Traceability is becoming table stakes. Meaning is the harder problem. Sponsors can already show a regulator where a number came from. What they cannot easily show is that the number was collected under the definition the analysis assumes, rather than against a superseded protocol version or a different population. That check belongs where the data is assembled, not in the document written at the end, which is the layer we sell.
A second wave of trials is forming outside the US
Its 2026-2030 industrial plan targets 20% annual growth and a quarter of the world's first-in-class drugs originating there 21.
Licensing of China-developed drugs grew 36% year over year, past $120B in total deal value 21. The buyers are US and European pharma.
A licensed-in asset still needs registrational trials that meet Part 11 and E6(R3), with every value traceable to its source, wherever it was collected.
Where we fit: more programs crossing borders means more trials that have to prove where each data point came from, across sites and systems a sponsor does not control. That is the layer we sell, and it is neutral to the country, the site and the trial database.
One layer between every data source and the sponsor's database.
It stores nothing. It checks study documents before enrollment, then cleans and delivers data for the whole trial.
What it is worth on one Phase III trial.
A made-up study, with the assumptions stated. 2,000 patients, 120 sites, 14 data sources, 3 years, a trial budget of about $82M at the published median of ~$41,000 per patient. Nothing here is a customer result: a measured before-and-after case study is the next milestone as the contracted trials run.
A hypothetical scenario, not measured performance. Published per-study figures scaled to 2,000 patients; the 90-day cleanup (against an industry average of 36 days 3) and the 60% query reduction are modelling assumptions. Direct cost and forgone sales are priced per day 7 and shown separately in this scenario. Depending on trial size, the model produces 63-80 days of earlier completion. The four numbers that land hardest on customer calls are enrollment start, query volume, earlier completion and the cost difference. These are the figures the contracted trials are being measured against; what is signed and live today is in the next section.
Six months and under $70K of total burn.
Everything above this line is modelled, except the back-test below. Everything below it is signed, audited or live.
Back-tested on 718 completed neurology trials (181,302 participants). On 260 held-out trial arms, 82% of outcomes landed inside the 80% forecast range, against 51% for the standard method, with 1.7% false alarms.
DIA invited the CEO to speak with a booth, and DIA's CEO offered a personal introduction to Pfizer. Nightingale is also a case study in USC's Regulatory Science curriculum.
Buyer access · meetings, reported as meetings
A $20B+ a year market for trial data, growing into much larger ones.
About 20,000 industry trials start each year 13. Running the whole data layer of one trial is worth roughly $1M, so the market for the product we sell today is $20B+ a year. Early contracts are smaller because they cover part of a trial; the same platform extends to adjacent products with far larger markets behind them.
This is the market this brief is about. Every interventional trial has to get data out of hospital systems, labs and devices and into a study database a regulator will accept. That work is bought today, from CROs and data-management teams, on every study. We replace the manual part of it.
What one trial is worth at full scope: Phase III $1.2-1.4M, Phase II ~$550K, device ~$450K. Contracts start smaller, at the rulebook and gap analysis, and grow as the pipeline runs for the life of the study. The second trial at the same sponsor costs us far less to deliver, because the connections and the format library already exist.
The pool keeps growing: more than 35,000 new studies register each year, about 20,000 of them addressable 13. Layers 2 and 3 below are the same engine pointed at adjacent work. They are upside, not the plan.
About 15,000 prospective studies a year sit in this layer, the balance of the 35,000+ registered annually once the ~20,000 in-scope interventional trials are counted 13. They are the prospective studies that run after approval: Phase IV and post-marketing commitments, prospective real-world evidence studies, device post-market surveillance, and diagnostic and digital-health studies. They pull from the same sources under the same rules, and run for years rather than months. Behind them sits pharma R&D of about $300B a year, plus devices, diagnostics and digital health at about $50B 12.
Once many trials have been run through the same checks, their data shares one structure. That structure is a scaffold: a frame with the shape of a real trial, which can be filled with modelled patients instead of recruited ones. Regulators already accept a limited version of this, called a modelled or synthetic control arm, where the comparison group is generated from prior trial data rather than by giving patients a placebo. Further out, the same scaffold supports simulating a candidate drug before it is tested. This layer is not a product today and is not in any forecast 11.
The trial pool this applies to
In scope means interventional trials collecting from more than one electronic source and reporting into a study database. Registry-only and single-source observational studies are excluded. Industry-sponsored trials carry full pricing; academic and hospital studies are served at smaller scope, which is why the $20B+ figure is an upper bound rather than a forecast.
Longer term: the same data supports modeled control arms, which regulators already accept in steps (FDA 2023 guidance, an EMA-qualified digital-twin method 20).
Sponsors pay per trial. CROs distribute. Margins look like software.
Sponsors hold the budget and pay per trial. CROs bring portfolios of them. We set the payment structure with each sponsor and break it across the trial rather than asking for it up front. Every engagement starts with the rulebook and gap analysis, and the aim is a sponsor who signs for the whole product, the rulebook and gap analysis plus the data pipeline, in which case the rulebook and gap analysis are free. Price a study in the calculator below.
Start of enrollment, midpoint, end. It matches the sponsor's own spend curve and shortens the procurement argument.
The rulebook and gap analysis is the fast, low-risk way in, and we give it away when the sponsor commits to the full product. We trade margin at the entry point for a place in the study, which is where the pipeline earns for the life of the trial.
In the modelled Phase III, direct savings of about 1.6x the fee from fewer late days and less data-handling work. Earlier sales, if the drug is approved, are counted separately.
Go-to-market
How sponsors buy · price a trial
Discount grows with each doubling of trials committed, capped at 10%.
Direct benefit over the trial
in direct savings for every dollar spent on Nightingale
Budget from ~$41,000 per patient 10; savings from a data-management team scaled to enrollment 89 and delay days at $55,716 direct cost per day 7. Earlier revenue, at ~$500,000 per day, is counted in full only for studies that lead to market (15% at Phase I, 35% at Phase II) and is shown separately: it is not in the total or the return. Days saved is an output, not an input: 14 days of head start before the first patient, plus up to 49 days of cleanup slip avoided, 8 days per connected eSource.
Why incumbents cannot easily copy this.
EDC vendors are not neutral to competing EDCs or to each site's own systems.
Each trial adds connectors, and CRO partnerships are exclusive.
Every trial adds lab and device formats that a single sponsor cannot rebuild.
Validated, source-linked data across sponsors.
Target: 36+ live trials by August 2027.
Top-10 embedded: 60 trials/yr at $2M = $120M.
Targets, not commitments. The ramp requires pipeline conversion above 35%.
Three full-time founders, each with a prior exit.
First hire: a commercial lead with clinical-trial domain fluency. Selling is learnable; knowing how a sponsor actually buys, and how a study runs, is the scarce part in this market.
Advisors: a sitting FDA advisory committee member, USC Regulatory Science faculty, senior sponsor clinical-ops leaders, and two founders who exited a trials company to a major Japanese CRO. All in a personal capacity; no institution endorses Nightingale.
Raising $2.5M on a SAFE for 24 months of runway.
Allocation follows our financial model, scaled to $2.5M with a 2% contingency. Team grows from 4 to 7 full-time. Fixed cost and break-even are estimates at full team; full model available in the data room.
Return potential · comparable revenue multiples 18
$100M at 25-50× is a $2.5-5B outcome. The billion-dollar line is crossed at $20-40M of revenue.
Comparables are illustrative, not a forecast. A seed valuation prices the outcome; see §12 for what could go wrong.
Key risks and mitigations.
Why
Emory's metrics are n=1.
Mitigation
The two contracted device trials are the test, measured identically. Each trial makes us integral to that sponsor's build.
Why
The ramp needs 35-50% of 37 trials; no conversion history yet.
Mitigation
Rulebook-check entry lowers the first yes; CRO partners convert portfolios. First metric at every board meeting.
Why
88% of contracted value came through one partner.
Mitigation
Exclusivity means a partner's full trial flow comes with it; two more partners targeted through DIA; direct sponsor conversations in parallel.
Why
$340K contracted, $40K being paid out. The two device trials begin when their sponsors close financing; pharma pays on 60-90 day terms.
Mitigation
Buffer carried in the raise; earlier milestones on new contracts; the raise does not depend on either trial's start date.
Why
We touch patient data.
Mitigation
GxP complete, SOC 2 in progress, cyber and E&O insurance; storage stays with the sponsor, limiting what any incident can reach.
Why
Relationships in the CEO; architecture in the CTO.
Mitigation
Standard vesting; the model and pipeline are documented as validated systems, which the compliance posture requires anyway.
Investor FAQ
EDC vendors sell a database, so a layer that feeds every competing EDC works against their product. The sources also sit with each site: a 40-site trial means 40 different setups to connect, which a database vendor does not do.
Emory already has. The GxP vendor audit is complete, Part 11 alignment is in place, SOC 2 Type II is in progress, and an FDA advisory committee member is on the team.
Revenue follows a sponsor's pipeline. A top-ten sponsor runs hundreds of trials a year, and later trials reuse integrations already built. Exclusive CRO partners bring their trial flow with them.
The product we sell today addresses $20B+ a year: about 20,000 trials a year, each worth roughly $1M of data work at full scope. The same platform extends to health R&D more broadly and to modeled control arms.
Every agreement includes consented, de-identified derived-data rights, and the rulebook-check model is ours. Where a sponsor keeps its data, it still needs the structuring, linking and validation work we do.
Clinical trial data infrastructure, live and contracted.
Raising $2.5M on a SAFE. Terms on request.
Every study that closes sooner is a treatment that reaches the people waiting for it sooner. That is the reason to build this well, and the reason to build it now.
Talk to the founder.
Pick a time for an intro call, or reach out directly.
We will walk through the product, traction, and the round, and answer anything this brief leaves open.
Book an intro callWhat happens next
Product, traction and the $2.5M SAFE round.
Terms and any materials you request.
A call with our flagship deployment, on request.
Show all 20 sources
- Tufts Center for the Study of Drug Development (2021): Phase III protocols average 3.6 million data points, three times the volume of ten years earlier.
- Marra C. et al. (2020), npj Digital Medicine: connected digital products in trials grew ~34% per year, 2000-2017.
- Tufts CSDD and Veeva Systems, eClinical Landscape Study (2017): 257 organizations; median 8.1 days from patient visit to EDC entry; 36 days on average from last patient last visit to database lock.
- Advarra, Study Activation Survey 2023: N>500 sites; ~69% use six or more sponsor systems.
- Garza M. et al. (2026): manual transcription error rate 8.23%; integrated capture showed no detected transcription errors.
- Stokman P. et al. (2021), Journal of the Society for Clinical Data Management 1(1), DOI 10.47912/jscdm.20: 20 Phase III studies, 1,939,606 queries; ~96,980 per study; 42% resulted in a data change.
- Smith Z., DiMasi J., Getz K. (2024), Therapeutic Innovation & Regulatory Science 58(5):855-862: direct cost of trial delay $55,716/day; opportunity cost on the order of $500,000/day.
- Salary.com, US Clinical Data Manager compensation, September 2026: average $144,625, ~$70/hour.
- Vendor-sponsored industry survey, n=88: ~12 hours/week per data manager per study on query and reconciliation work. Indicative.
- Moore T. et al. (2018), JAMA Internal Medicine: median cost per patient in pivotal trials ~$41,117. Per-source and fixed-cost components are estimates.
- IQVIA Institute, Global Use of Medicines: global medicine spending ~$1.6T. Medical device market ~$600B (industry estimate). WHO: global health spending ~$10T.
- Evaluate Pharma / IFPMA: pharmaceutical R&D ~$300B per year; medtech and digital-health R&D ~$50B (estimate). IQVIA Technologies, Veeva and Medidata combined revenue ~$10B.
- ClinicalTrials.gov: more than 35,000 new studies registered per year. The ~20,000 addressable interventional figure is an estimate.
- ClinicalTrials.gov sponsor records: Pfizer ~423 tracked / 251 active; Eli Lilly ~333. Order-of-magnitude proxy.
- ICH E6(R3): §3.16.1(c), §3.16.2(c), §2.12. Step 4 6 Jan 2025; EU 23 Jul 2025; FDA 9 Sep 2025; Annex 2 adopted 3 Jun 2026, effective 15 Jan 2027.
- FDA Digital Health Technologies guidance (Dec 2023); FDA Part 11 Q&A (Oct 2024); FDA and EMA AI principles (Jan 2026); FDA Real-Time Clinical Trials (28 Apr 2026); HHS and ARPA-H SURPASS (Simulation-augmented, Real-time Platform Adaptive Seamless Trials), solicitation ARPA-H-SOL-26-164 (30 Sep 2026).
- Disclosed transactions: Change Healthcare/Optum $13B; Datavant/Ciox ~$7B; Komodo Health $3.3B; Flatiron Health/Roche $1.9B; Truveta >$1B; Tempus AI market capitalization ~$10.6B.
- Public revenue multiples: Snowflake at IPO >100× trailing; Palantir ~80× (2025); Cloudflare ~35× (2025); Databricks ~27× (private, 2025).
- Private comparables: a September 2026 software seed round at $7M on $100M post-money with ~$500K lifetime revenue; a clinical-trial software company that raised under $400K in total and sold for roughly half a billion dollars. Details available on request.
- FDA, Considerations for the Design and Conduct of Externally Controlled Trials, draft guidance, February 2023. EMA, qualification opinion on prognostic covariate adjustment using digital twins (PROCOVA), September 2022.
- BioSpace, "China plans to transform biotech sector into $522B powerhouse by 2030," September 18, 2026, reporting China's Ministry of Industry and Information Technology 2026-2030 plan: 20% annual industry growth, 25% of global first-in-class drugs originating in China, and licensing of China-developed drugs up 36% year over year, past $120B in total value.
- Assessment of Trends in the Design, Accrual, and Completion of Trials Registered in ClinicalTrials.gov by Sponsor Type, 2000-2019, JAMA Network Open (2020): of 245,999 interventional studies, 70,329 (28.5%) were industry-sponsored.
Plain-language glossary
No matching terms.