GLP study report writing from raw data, with every number traced to its source cell
Raw study data becomes a draft report in hours, with every number traced to its source.
See one case, screen by screen ↓The study is over and the data package has landed: the signed study plan, a dozen raw data tables — body weight, food consumption, clinical signs, clinical pathology, organ weights, macroscopic and microscopic findings, toxicokinetics — and a long report template. Someone now has to turn thousands of values into report tables and results text, group by group and sex by sex.
Most of that time is re-keying and checking. Every table is typed again, every mean and percent change is copied into the text, and every number is then checked back against the raw data. A missing table the plan requires surfaces late, often in review. And one transcribed value that does not match its source is a finding at the Quality Assurance audit.
Estimated split for a typical repeat-dose toxicity report, by hand and with the solution.
Six specialist agents read, build, draft and check every report; a study director reviews and e-signs.
VLT-TOX-028 is a 28-day subcutaneous toxicity study of Veltrimab in Sprague Dawley rats, with a 14-day recovery period. Its data package arrived on 05 Oct 2026; the report is due Oct 16. Here is what happens next, screen by screen, in the working solution.
Lena opens the solution: 3 reports in scientist review, 3 data packages ready to draft, 1 open gap and 4 reports signed in the last 30 days. VLT-TOX-028 sits at the top of her list — 13 tables received, ready to draft. Two other reports are already being drafted.
Consistency checker: “1,204 numbers re-checked today · 0 mismatches”
Four inputs, each marked received: the study plan v2.0 with amendment 1 (18 pages, signed 04 Aug 2026), the raw data package — 13 tables and 3,842 values across BW, FC, CL, LB, OM, MA, MI and TK — the repeat-dose report template v5.2 with its 9 sections, and the toxicokinetics contributing report signed by Sam Patel on 01 Oct 2026.
The study plan reader finds 4 groups, 100 animals plus 42 for toxicokinetics, and 9 report sections required. The raw data reader loads all 13 tables. The table builder fills the report tables straight from the source cells, section drafters write all 9 sections in parallel, and the consistency checker compares 152 numbers with their source — 0 mismatches.
“Report tables built straight from source cells — nothing re-keyed”
9 sections drafted, 152 numbers traced to source, 0 tables re-keyed, 2 gaps to resolve. The gap finder puts them first: recovery-group histopathology is missing from section 4.5, which blocks signature, and one food consumption value was excluded in section 4.2, which needs her confirmation.
In males at 100 mg/kg/week, body weight gain over Days 1–29 was 142.7 g against 156.8 g in controls (−9.0%; p<0.05), with food consumption of 25.3 against 26.6 g/animal/day. Hovering any number shows the table, row, column and cell it came from; a percent change shows both cells and the formula.
Source cell: Food consumption — males (g/animal/day) · Mean Weeks 1–4 · Group 1 · control · FC_Males · B6
Clicking 142.7 opens Body weight — males (g) from VLT-TOX-028_BW.xlsx, with group means, SDs and significance markers, and the cell itself highlighted: Gain Day 1–29 · Group 4 · 100 mg/kg/week = 142.7, BW_Males cell E7.
“Raw data is read-only. The agents copy the cell reference, never the typed value.”
Week 3 food consumption for cage M-4-07 (Group 4 males) was left out of the group mean. The raw data note says why: “Food spilled — weighing invalid.” Lena confirms the exclusion, and the section drafter adds a note under Table 4-1 saying so, and that it did not affect the interpretation.
The gap quotes its evidence. Study plan v2.0, section 9.4, requires microscopic examination of Groups 1 and 4 “at the terminal and recovery necropsies”. The data package manifest lists MI_Terminal.xlsx, but MI_Recovery.xlsx is not present — so sections 4.5, 1 and 5 cannot state whether the liver findings are reversible.
Two choices on the gap: “Request from pathology” or “Add a statement” — a pending-data statement in the report.
Lena chooses “Request from pathology”. Dr. Omar Haddad provides MI_Recovery.xlsx — 20 animals, signed 06 Oct 2026. The section drafter rewrites 4.5 and updates the summary and conclusion; the new numbers are traced to the recovery table with 0 mismatches. After 14 days without treatment, no hepatocellular hypertrophy remained at 100 mg/kg/week (0/5 males, 0/5 females).
Lena approves all 9 sections and signs version 1.0: 158 numbers traced to source, no open gaps. She enters her password and picks the meaning of the signature. The signature is linked to that exact version, the report is locked, and it goes to Dr. Maya Chen for the Quality Assurance audit.
“Signed by Dr. Lena Ortiz, Study Director · 06 Oct 2026 · 09:23 UTC” · Meaning: “Study director approval — release to Quality Assurance for audit”
The same report, seen by the people who carry it — what their week looked like, and what it looks like now.
Six specialist agents read, build, draft and check every report; a study director reviews and e-signs.
Reads the signed study plan and amendments: groups, doses, animal numbers, endpoints, key dates and the report sections the template needs.
Loads the raw data tables — body weight, food, clinical signs, laboratory, organ weights, macroscopic and microscopic findings, toxicokinetics — maps each to its domain and checks animal counts, days and groups against the plan.
Builds every report table cell by cell from the source tables, with significance markers — nothing is typed by hand.
Writes each report section in house style, describing test item-related changes by group and sex with direction, size and significance. Every number is inserted as a reference to its source cell.
Compares what the study plan requires with what the data package contains and what the template needs, and lists every missing or unsigned item, empty template field and exclusion that needs a statement.
Re-reads every number in the draft against its source cell, checks that the summary and conclusion repeat the results exactly, and that percent changes are calculated from the cited cells.
Reviews, edits and e-signs. The agents propose; a named person decides.
Study directors can ask about any study or report — or tell it what to change. Every number in an answer links to its source cell.
Summarise the 100 mg/kg/week findings in VLT-TOX-028
Male body weight gain 142.7 g vs 156.8 g in controls (−9.0%); ALT 53 U/L vs 38 U/L (+39%); liver weight 14.21 g vs 12.84 g; centrilobular hepatocellular hypertrophy in 4 of 10 males, minimal to mild; globulin up to 3.4 g/dL — the administered antibody. None looks adverse. With the recovery data in, all findings showed full or partial reversal after the 14-day recovery period.
What is blocking signature on VLT-TOX-028?
One blocking gap: recovery-group histopathology is missing. The study plan (section 9.4) requires microscopic examination of Groups 1R and 4R, and the data package has no MI_Recovery file — so sections 4.5, 1 and 5 cannot state reversibility. There is also one minor item: a spilled food weighing to confirm.
Add a rule: highlight any group mean more than 15% from control
Done. Every traced number in a draft that differs from its control by more than 15% is marked in amber — in VLT-TOX-028 that is ALT (+39%) in 4.3, globulin (+19%, +31%) in the Summary and relative liver weight (+15%) in 4.4. You can change the threshold in Settings → Rules.
Show the reports waiting for my review
Three reports are in Scientist review: CRV-TOX-014 (due Oct 9, 1 gap), NXL-TOX-011 (due Oct 12) and CRV-GTX-007 (due Oct 10). I have filtered the study list for you.
13 screens from the working solution, on its sample data. Pick one to see it large.
Templates, number style, reviewers and the rules checked before signature are settings, not a project.
| Setting | Default | Choose from |
|---|---|---|
| Report template | Repeat-dose toxicity report v5.2 | Repeat-dose toxicity report v5.2 · Repeat-dose toxicity report v5.1 · Genetic toxicology report v3.1 · Safety pharmacology report v2.4 |
| Number style | Mean to 1 decimal, SD in brackets | Mean to 1 decimal, SD in brackets · Mean ± SD · Mean to source precision |
| Significance markers | * p<0.05 ** p<0.01 | * p<0.05 ** p<0.01 · a p<0.05 b p<0.01 · None |
| Pathology terms | Standard nomenclature (INHAND) | Standard nomenclature (INHAND) · As recorded by the pathologist |
| Histopathology section reviewed before the study director by | Dr. Omar Haddad | Dr. Omar Haddad · Dr. Maya Chen · Sam Patel |
| Block signature while a blocking gap is open | On | On · Off |
| Draft automatically when a data package is complete | Off — the study director starts drafting | On · Off |
| Highlight differences from control above | Off — threshold 15% | On · Off · threshold 10 · 15 · 20 · 25 % |
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 study data · “target” = the design goal, measured in the live solution; “by hand” = the same team’s reports before agents · “estimated” = our estimate. People, products and studies named on this page are characters in the working solution.
A solution that turns a GLP study’s data package — the signed study plan, the raw data tables and your report template — into a draft study report. Six agents read, build, draft and check it, with every number linked to its source cell; the study director reviews, edits and e-signs.
Nothing is re-keyed. The table builder fills every report table from the source cells, the section drafter inserts each number as a reference to its cell, and the consistency checker re-reads every number against its source before signature. A mismatch blocks signature.
No. The section drafter describes the changes by group and sex and proposes adversity and the NOAEL; both are marked for the study director, who decides. Agents draft and check — they never sign.
The gap finder lists it with the study plan clause and the data evidence, and classifies it as blocking or minor. A blocking gap stops signature until it is resolved — for example by requesting the data — or stated in the report. When new data arrives, only the sections it feeds are redrafted.
No. Raw data is read-only for the agents and for people. Excluded values stay in the raw data and are recorded with their reason, and the report states the exclusion once the study director confirms it.
The working solution drafts repeat-dose, genetic toxicology, safety pharmacology, reproductive toxicology, toxicokinetic and acute studies, each with its own section template. Your templates, number style, significance markers and pathology terms are settings.
The study director signs with a password and a stated meaning, such as “Study director approval — release to Quality Assurance for audit”. The signature carries the date and time, is linked to the exact version, and the signed draft goes to your Quality Assurance auditor automatically.
The Agentic Solution Engine builds and deploys it from your requirements — your report templates, house style and a sample of completed studies — and it goes live once every quality gate has passed. We will walk you through it on one of your own studies first.
Ask the literature in plain words and get a sourced answer in seconds.
1-min filmPipeline WatchEvery rival trial, readout and approval lands in a sourced brief for each programme, every Monday.
1-min filmScience HarvesterEvery inherited file read and cited, and a signed dossier for each asset before the experts leave.
1-min filmWe’ll draft one of your own completed studies, from its data package to a report ready for your signature.