Batch record review by exception: every page read by agents, only the exceptions for QA
Every executed batch record read page by page, so QA opens only the entries that break a rule, each linked to its page.
See one case, screen by screen ↓An executed batch record for one batch of tablets runs to 300 pages or more: dispensing, granulation, drying, compression, coating, packaging, the reconciliation. Before the batch can go to disposition, a QA reviewer reads every one of them — every value against its limit, every “Done by” and “Checked by”, every correction, every date and time.
Almost all of those pages are clean. The few problems that matter hide in the gaps between pages: a hold time that only shows when you put the granulation end on page 58 next to the drying start on page 81, a quantity that differs between the step and the reconciliation 246 pages later, an overwrite with no initials. Line-by-line review finds them by reading everything, slowly — and the release date waits.
Estimated split of the 31 h line-by-line review and the ≈10.8 h target by exception, for one paper record.
Six agents read every page, check every entry and write each exception; the QA reviewer decides and signs.
CR26-1108 — 312 paper pages for a batch of Corventa 20 mg film-coated tablets made at Plant 2, Westbrook, with a release target of Oct 9. Here is its review, screen by screen, in the working solution.
Dana opens the review bench for three plants: 2,115 pages read on six records, 34,589 entries checked, 22 exceptions raised, 15 that need her — 2 of them critical. Each record is drawn as one bar per page, with a mark wherever there is an exception.
“You open 1 page in 85 · Clean pages spot-checked at 2 %”
CR26-1109 came in from the Plant 2 scan station: 304 of 304 pages of the issued copy present. The agents read 41 handwritten pages with 0 illegible boxes, extract 4,270 entries, check 188 master-record steps and 14 hold times, and list 4 exceptions — the first a final blend time of 12 minutes against 15 ± 1 min. 300 pages come out clean.
“Six agents read every page before you open it.”
On CR26-1108, granulation ended on Oct 1 at 21:40 (page 58) and drying started on Oct 2 at 23:55 (page 81): the wet granule was held 26 h 15 min. Step 6.4 of master batch record MBR-CRV-020 v7.2 allows 24 hours and calls hold time a critical process parameter. The drying page records “N/A” for a deviation, so none was raised. Both pages sit side by side, with the calculation between them.
Suggested: raise a deviation — “Dissolution and bioburden impact must be assessed before disposition.”
The exception writer has pre-filled DEV-26-0631: batch, title, the evidence on p.58 → p.81 and master record step 6.4, classification “Critical — CPP outside limit”, owner Sam Patel in production with Dana for QA, batch on hold. Dana checks it and raises it; because it is critical, Omar Haddad, the QA operations lead, second-signs.
Page 21: the API was weighed at 12.004 kg, inside 12.000 ± 0.010 kg, but the “Checked by” box is empty — and the master record makes this a second-person verification. The suggestion is a production query, not a signature.
“A check cannot be signed afterwards — if nobody witnessed it, raise a deviation instead.”
Step 6.3 on page 52 records 41.6 kg of purified water for the binder; the material reconciliation on page 298 shows 46.1 kg used and nothing returned — 4.5 kg apart. Coating spray on page 181 started at 13:10, while compression on page 174 ended at 14:25. Loss on drying of 2.9 % was re-dried to 1.8 % as step 7.6 allows, so Dana accepts that one with a reason.
Nine exceptions decided: one deviation, one accepted, seven corrections — the missing check, the water quantity, the coating time, an overwritten tablet weight, a blank line clearance field, a yield recorded as 98.4 % that the entries give as 96.1 %, and a late entry not marked. They go to Sam Patel, production supervisor at Westbrook, on one list, due Oct 8 at 16:00.
“Rescan only the corrected pages — the agents re-read those pages.”
Sam returns 9 rescanned pages. The page reader matches them to the issued copy, the entry extractor reads the 7 corrected entries with the originals still legible, the GDP checker confirms each one is initialled, dated and has a reason, and the reconciler confirms times, quantities and yields are consistent.
Dana opens a random 2 % sample of the 300 clean pages — 6 pages, every entry already checked by the agents. If she finds something, “I found something” raises it as an exception and logs a miss against the validation, and the validation lead is notified. This time: clean.
The summary reads: 9 decided — 1 deviation, 7 corrected, 1 accepted; 300 clean pages, 6 spot-checked, no misses; DEV-26-0631 open. Dana signs with her password and the meaning “Production record review complete — 21 CFR 211.192”, and the record goes to Dr. Maya Chen, Qualified Person, with the deviation attached.
“Your name, the time and the meaning are bound to this record and shown on every copy. Agents cannot sign.”
The same record, seen by the five people who carry it to release — what their week looked like, and what it looks like now.
Six agents read every page, check every entry and write each exception; the QA reviewer decides and signs.
Receives scanned, hybrid and eBR records, checks the page count against the issued copy, classifies every page and reads print and handwriting.
Turns every box on every page into a typed entry — value, unit, initials, signature, date and time — with page and position, and marks overwrites, strike-throughs, blanks and N/A.
Checks each entry against the master batch record and specifications — limits, critical process parameters, required steps and checks — and follows the retest and rework paths the master record allows before raising an exception.
Applies the good documentation practice rules: signatures, second-person checks, corrections, blanks, late entries, date formats, and initials against the signature roster and training records.
Builds the batch timeline and material balance: hold times between paired steps, step order, every yield and reconciliation recomputed, each quantity compared with every other mention of it.
Merges duplicate findings and writes one plain exception per problem, with page citations, the rule, severity, confidence and a suggested decision — raise a deviation, send a correction, or accept.
Decides and signs. The agents propose; a named person decides.
Ask about any record, exception or rule — or say what to change. Every answer points to its page.
What needs me on CR26-1108?
9 open on CR26-1108 (Corventa 20 mg, 312 pages; 300 clean). First, critical — wet granule held 26 h 15 min, limit 24 h: granulation ended Oct 1 21:40, drying started Oct 2 23:55; step 6.4 makes hold time a CPP. Suggested: raise a deviation. Then the majors, the GDP corrections for Sam Patel, and one check — loss on drying re-dried as step 7.6 allows.
Which records can I sign today?
One now, and one more after a single decision. VL26-0428 came through clean — 478 eBR pages, nothing raised; spot-check 10 pages and sign. VL26-0427 needs one decision: buffer BF-12 used at 74 h against a 72 h expiry.
Flag any value within 5 % of its limit
Done — added as a check-only rule. On the open records it finds 1: tablet hardness 8.3 kP at 07:00 on Oct 3, CR26-1108 p.147, inside 8.0–12.0 kP — the press was adjusted and the next check was 9.6 kP. Saved as rule set v3.2; check-only rules add exceptions and never hide one, and Lena Ortiz signs the validation addendum.
How do I know the clean pages are clean?
Every record gets a random spot-check of 2 % of its clean pages. Last month reviewers opened 298 clean pages and found 0 misses, and reviewers kept the agents’ suggestions 91 % of the time. The one miss this quarter, an unmarked late entry, became a rule in v3.1.
13 screens from the working solution, on its sample data. Pick one to see it large.
What reaches a person, who signs and who corrects are settings — each change versioned and approved.
| Setting | Default | Choose from |
|---|---|---|
| Steps with critical process parameters | Always shown to the reviewer | On · Off |
| Values close to a limit | Off | On (raise a “check” within 5 % of a limit) · Off |
| Late entries not marked “late entry” | Raise an exception | On · Off |
| Spot-check of clean pages | 2 % of clean pages | 2 · 5 · 10 % |
| Record types in scope | Paper, hybrid, electronic | Paper · Hybrid · Electronic |
| Second signature on critical exceptions | On — the QA operations lead signs after the reviewer’s decision | On · Off |
| Deviation owner | Production supervisor of the plant | Production supervisor · QA operations lead |
| Signature meaning | Production record review complete — 21 CFR 211.192 | Production record review complete — 21 CFR 211.192 · Batch record reviewed — EU GMP Chapter 4 |
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 plant data · “target” = the design goal, measured in the live solution · “estimated” = our estimate · “typical” = published figures (FDA 483 observations for sterile sites, 2024 — production record review, 21 CFR 211.192, the third most cited). People, companies and products named on this page are fictional — characters and sample data in the working solution.
Instead of a QA reviewer reading every page of an executed batch record, every page is read and checked first, and the reviewer sees only the entries that break a limit, miss a signature or contradict another page — each linked to its page. Here, six agents do the reading and checking; the QA reviewer decides each exception and signs the review.
Yes. The same agents read paper scans, hybrid bundles and eBR exports. On eBR records they work through the event log too: alarms acknowledged without a comment, overrides of an eBR check, and e-signatures by people not yet trained on the current procedure.
The cross-page reconciler builds the batch timeline and material balance from every entry. It computes hold times between paired steps, checks step order, recomputes yields and reconciliations, and compares each quantity with every other mention of it — then shows both pages and the calculation.
No. The agents suggest raising a deviation, sending a correction or accepting; a named QA reviewer decides each exception and signs the review. Critical exceptions need a second signature from the QA operations lead, and agents cannot sign.
Every record gets a random spot-check of its clean pages — 2 % by default — before it can be signed. A problem found there is raised as an exception and logged as a miss against the validation, and spot-check misses and reviewer agreement are measured every month.
Corrections go to the plant’s production supervisor on one list, with the page, what to correct and the rule. Production corrects by your GDP SOP and rescans only the corrected pages; the agents re-read those pages and confirm each correction before the reviewer signs.
Yes. Which steps always reach a person, near-limit checks, late-entry rules, the spot-check rate, record types in scope, second signatures, deviation owners, who receives corrections and the signature meaning are all settings. Each change is a new rule-set version approved by your validation lead.
The Agentic Solution Engine builds and deploys it from your requirements and documents — your master batch records, specifications, GDP SOP and a sample of executed records — and it goes live once every quality gate has passed. We will walk you through it on your own records first.
Any batch-record question answered in seconds, cited to the exact page and step.
1-min filmValidation DocumentsValidation protocols and reports drafted in minutes, every value traced to its source page.
1-min filmAsk the PlantBatch analysis that took three months now takes hours, with the query behind every answer.
1-min filmWe’ll run Batch Record Review on a sample of your own executed records.