Investigator-initiated study management that shows first whether an ongoing study already answers the question
Proposals decided in half the time with a cited review packet.
See one case, screen by screen ↓Software that takes an investigator-initiated study (IIS, also called investigator-sponsored research) from the concept an investigator submits to the committee's decision and the letter. Investigator-Initiated Study Review adds seven agents that score each concept on priority fit, scientific quality, feasibility and overlap, rank the cycle on one scorecard and build a cited packet for the scientific reviewer to sign and the committee to decide.
How to choose AI solutions for life sciences →
Updated
Every concept that comes through the portal has to be judged the same way: does it fit a published research priority, is the design sound, can the site actually enrol, is the budget at fair market value, are there compliance red flags — and does a company study, another funded investigator study or a registered trial already ask the same question. The IIS programme manager chases what is missing, a medical director reviews it on top of a full medical role, and the committee meets once a month.
The step most often missed is overlap. Reviewers check it from memory, so a concept that repeats the company's own phase 3b — same population, same endpoint, same timepoint — can reach the committee, cost a slot and a reviewer's afternoon, or be funded as a duplicate. Feasibility is taken as written, budgets are compared with benchmarks line by line in spreadsheets, and the compliance check depends on someone reading the cover email.
Median days per step on the working solution's dashboard: by hand in 2025 (52 days) and with the agents in the Q4 2026 cycle (27 days).
Seven specialist agents read, score, search for overlap, cost and screen every concept and rank the cycle; the scientific reviewer confirms and signs, and the committee decides.
VELOCITY-CD — Dr. Elena Varga at Danube University Hospital, Budapest, proposes to study early endoscopic response to Veltrimab after anti-TNF failure in Crohn's disease. It arrives on Oct 6 for the Q4 2026 cycle. Here is what happens next, screen by screen, in the working solution.
Maya opens the Q4 2026 scorecard: 10 concepts open, 4 decided, committee on Wed Oct 21. Each concept is scored on priority fit, scientific quality, feasibility and overlap, weighted 30 · 30 · 20 · 20, and ranked out of 100. SCALP-V leads at 85; TYK-CLE sits below the line, not ranked until compliance clears it. VELOCITY-CD is new and not scored: a 6-page concept, a $448,000 budget at 30% overhead and a cover email.
“Score = weighted mean of the four criteria, out of 100 · the scientific reviewer confirms every score.”
The concept reader turns the concept form, budget sheet and cover email into 44 fields, each cited. The priority matcher finds IMM-3 at a fit of 0.88 — 4.4 of 5. The design appraiser gives 3.9: single arm, validated endpoint, power calculation missing. The feasibility and budget checker gives 2.4. Before the overlap check, VELOCITY-CD scores 74 and sits seventh.
Feasibility and budget checker: “1.1 patients a month at this site → 55 months.”
The overlap finder searches 96 company studies, 61 funded investigator studies and 255 public registry trials. VELTRIS-CD, the company's registered phase 3b (EU CT 2024-512386-41-00), has the same population, treatment, primary endpoint — SES-CD decrease of at least 50% — and week-12 timepoint, reading out in Q2 2027. Overlap 0.86. The score falls from 74 to 62, last of nine.
“What VELOCITY-CD adds: transmural healing on MR enterography at week 24 — the gap priority IMM-3 names.”
Maya opens the review packet. Seven agents built it; every step is listed with what it found. The structured record shows sponsor, product, design, population, primary endpoint, 60 patients in 18 months and the support requested — and what is missing: a power calculation and an ethics plan. Every field links to its passage in the concept form, open beside it.
The grid compares each study on population, treatment, control, outcome, timing and region. VELTRIS-CD scores 0.86, the funded study HISTO-CD 0.54, the public trial SONO-CD 0.38. What is the same cites the registry record; what the concept adds cites the concept. A timeline shows when each study answers its question.
“VELTRIS-CD already answers the primary question in Q2 2027. Danube University Hospital was one of its sites. Ask for transmural healing as the primary endpoint — the gap our priority names.”
The site enrolled 1.1 patients a month in VELTRIS-CD and in its registry study, with 35% screen failure — about 55 months for 60 patients. Maya tries the scenarios: add a second site, 23 months; 40 patients, 37 months; both, 15 months.
Each budget line is set against the 2026 benchmarks by country. Central endoscopy reading at $520 a read is above the $260–$340 range, and institutional overhead is 30% against a 25% cap. With both at benchmark, the budget falls from $448,000 to $397,750, inside the $360k–$410k range. Product for 60 patients over 52 weeks: 180 IV vials and 360 prefilled pens, which Sam Patel confirmed from clinical stock.
The checker's guardrail: “Never edits the investigator's budget.” The reviewer decides, and the letter asks for the change.
The packet writer drafted a cited summary with seven citations and proposed four conditions: transmural healing on MR enterography at week 24 as the primary endpoint; a second site, or 40 patients justified by a power calculation; overhead at 25% and central reading at the benchmark rate; a safety data exchange agreement before product is released. Maya recommends a full protocol with conditions and signs.
“Meaning of this signature: I reviewed this proposal and its cited evidence and make this recommendation to the IIS review committee.”
VELOCITY-CD joins NEXA-FRAIL, SCALP-V, NAIL-V and CORE-PsA on the docket. Each packet shows the reviewer's recommendation and summary, the criteria scores — for VELOCITY-CD, novelty 2.6 and feasibility 2.4 — the overlap and the budget flags. Six voting members, quorum four; drug supply and compliance sit in without a vote.
All six vote to request a full protocol, Sam Patel confirms stock and Dana Okafor has no objection. Dr. Omar Haddad signs the decision. The letter writer drafts the letter with the four numbered conditions and a due date of January 4, 2027; a policy check removes internal study names, and it waits for Lena Ortiz to send through the portal.
Policy check: “Internal study names removed · plain words · due date added.”
The same concept, seen by the five people who carry it from the portal to the letter — what their cycle looked like, and what it looks like now.
Seven specialist agents read, score, search for overlap, cost and screen every concept and rank the cycle; the scientific reviewer confirms and signs, and the committee decides.
Turns each concept form, budget sheet and cover email from the submission portal into a structured proposal record — every field with the passage it came from.
Matches the concept to the published research priorities and quotes the priority text it fits, with a fit score and the runner-up priorities.
Scores scientific quality 1–5 on the committee rubric — a clear question, a validated primary endpoint, a design that answers it, a power calculation — and names what is missing.
Searches company studies, funded investigator studies and the public trial registry, compares each candidate on population, intervention, comparator, outcome, timepoint and region, scores the overlap and names what the concept adds.
Projects months to full enrolment from the site's own past studies, checks every budget line against fair-market-value benchmarks and the overhead cap, and works out the drug quantity.
Screens the concept and correspondence for red flags — commercial colleagues involved, promotional wording, links to prescribing — and notes payments in the last 24 months, transparency and safety-agreement obligations.
Drafts the cited reviewer summary and the one-page committee packet, and after the decision drafts the letter to the investigator with the numbered conditions and the date the full protocol is due.
Recommends and e-signs. The agents propose; a named person decides.
Ask why a concept ranks where it does, whether it overlaps, whether the site can enrol, or what is wrong with its budget. Every answer points to the passage behind it.
Why is VELOCITY-CD ranked last?
It scores well on fit — 4.4 for priority IMM-3 — and 3.9 on design, but two things pull it down: an overlap of 0.86 with VELTRIS-CD, which has the same population, treatment, endpoint and week 12; and feasibility of 2.4, since 1.1 patients a month at this site means 55 months. Before the overlap check it scored 74; with it, 62. If the primary endpoint became transmural healing on MR enterography at week 24, the overlap would drop and it would rank near the top half.
Does VELOCITY-CD overlap with our own studies?
Yes — one study asks the same primary question. VELTRIS-CD, our phase 3b, scores 0.86: same population, endpoint and week 12, reading out in Q2 2027. HISTO-CD, a funded study, scores 0.54 and a registry trial in Germany 0.38. What VELOCITY-CD adds is transmural healing on MR enterography — exactly what priority IMM-3 asks for.
Can VELOCITY-CD enrol 60 patients in time?
Not at this site alone. 60 patients in 18 months needs 3.3 a month; Danube University Hospital enrolled 1.1 a month in VELTRIS-CD and in its registry study, with 35% screen failure. At this site's rate, 55 months; with a second site, 23; with 40 patients, 37; with both, 15. The concept has no power calculation yet — that is condition 2.
What is wrong with the VELOCITY-CD budget?
Two lines: central endoscopy reading at $520 a read against a benchmark of $260–$340, and overhead at 30% against the 25% cap. With both at benchmark the budget falls from $448,000 to $397,750, inside the fair-market-value range of $360k–$410k. Everything else is within range.
13 screens from the working solution, on its sample data. Pick one to see it large.
The rules the agents apply, who reviews and how the committee decides are settings — every change versioned in the audit trail.
| Setting | Default | Choose from |
|---|---|---|
| Scorecard weights | 30 · 30 · 20 · 20 | Priority fit · scientific quality · feasibility · overlap, adjusted on the scorecard |
| Overlap alert | 0.6 | 0.5 · 0.6 · 0.7 |
| Minimum priority fit | 0.6 | 0.5 · 0.6 · 0.7 — below it, a concept is outside the published priorities |
| Overhead cap | 25% of direct costs | Step up or down |
| Enrolment from site history | On | On · Off |
| Commercial-involvement screen | On — hold any concept whose correspondence names a sales or marketing colleague | On · Off |
| Budget benchmarks | 2026 benchmarks by country · v3 | 2026 benchmarks by country · v3 · 2025 benchmarks by country · v2 |
| Committee quorum | 4 of 6 voting members | 3 of 6 · 4 of 6 · 5 of 6 |
Every figure is labelled: a target the solution is built to, an estimate, a typical published result, or a proven one.
“demo” = seen in the working solution, on its sample concepts · “target” = the design goal, measured in the live solution (programme goal 40–50% faster review). Context on research support and transparency: OIG Compliance Program Guidance for Pharmaceutical Manufacturers (2003), CMS Open Payments, EFPIA Disclosure Code. People, institutions, products and studies named on this page are fictional — characters and sample data in the working solution.
Software that takes an investigator-initiated study (IIS, also called investigator-sponsored research) from the concept an investigator submits to the committee's decision and the letter. Investigator-Initiated Study Review adds seven agents that score each concept on priority fit, scientific quality, feasibility and overlap, rank the cycle on one scorecard and build a cited packet for the scientific reviewer to sign and the committee to decide.
The overlap finder searches company studies, funded investigator studies and public registry trials, and compares each candidate on population, intervention, comparator, outcome, timepoint and region. It scores the overlap, quotes both passages, shows when each study reads out and names what the concept adds. A study at or above the overlap alert — 0.6 by default — is marked high overlap.
Priority fit, scientific quality and feasibility are scored 1–5, and overlap from 0 to 1 counts against the concept — an overlap of 0.86 leaves 0.7 of 5. The score out of 100 is the weighted mean of the four. The committee's default weights are 30 · 30 · 20 · 20, and moving a weight re-ranks the cycle at once. The scientific reviewer confirms every score.
Yes. Every budget line is compared with the benchmark range for its country and cost item, and institutional overhead with the cap — 25% of direct costs by default. Lines above range are flagged with the passage from the budget sheet. The agent never edits the investigator's budget; the reviewer decides and the letter asks for the change.
The compliance screener reads the concept and correspondence for a sales or marketing colleague, promotional wording or a link to prescribing, and lists payments to the investigator in the last 24 months. A red flag holds the concept for compliance, scientific review pauses, and a held concept is not ranked.
No. Company studies are used to find overlap but are never shown to investigators. The letter writer never names internal studies, and a policy check removes internal study names before a letter is sent through the portal.
Yes. The agents score and draft and make no decisions. The scientific reviewer edits the summary, sets the conditions and signs a recommendation; the committee votes with a quorum; the chair signs the decision; compliance clears or holds red flags.
The Agentic Solution Engine builds and deploys it from your requirements and documents — your research priorities, committee rubric, budget benchmarks and a sample of past concepts — and it goes live once every quality gate has passed. We will walk you through it on your own concepts first.
Every claim checked against its reference before committee.
1-min demoMedical Information ResponseEvery sentence of the reply linked to its approved source, off-label parts evidence only.
1-min demoMedical Insights ManagementEvery field insight becomes a cited theme within days, with an owner for the next step.
1-min demoWe'll run Investigator-Initiated Study Review on a sample of your own past concepts and research priorities.