Batch Review & Release
Give every batch a full second check, in hours
Litewave applies your master record, your specifications and your release criteria to every operation in every batch, then gives QA a review-by-exception file with the evidence attached. Your reviewers keep the decision and get their time back. It runs against your eBPR and MES data, and against handwritten paper that most systems cannot read at all.
The review gets done. The question is what it costs you
Hours of concentration per batch, from your most senior reviewers, reconciling one record against four others.
A 100 to 150 page executed record holds thousands of individual entries. Every temperature reading, every duration calculation, every signature and every material charge has to be reconciled against the master record, the in-process specification and the analytical limits, which live in other documents and other systems. Your reviewers work through all of it. The cost is hours of sustained concentration per batch, from the most experienced people on site, under release pressure.
Sustained cross-document comparison is precisely the task where human attention degrades, and the entries that get missed are rarely the dramatic ones. A blank "checked by" on page 94. A duration off by ten minutes. A result inside specification on its own but drifting against the last twelve lots. Reviewer fatigue and fragmented source data, not competence, are what put those at risk, and every one of them is found by reading rather than by judgement.
The second cost is the queue. The same small group of senior reviewers is the bottleneck for disposition, for investigations and for everything else the site wants them on, which is why cycle times stretch into weeks and why adding a line means adding headcount.
Electronic batch records were supposed to solve this, and for high-volume single-product sites they did. For a contract manufacturer or a high-mix site the math inverts: every new product needs its own validated digital template, and many of those products will run a handful of batches before the contract winds down. The templating backlog grows faster than throughput. Litewave runs on top of the records your floor already produces, so a new product is live the day its SOP is.
Most systems need your records to be clean first
Litewave reads handwriting, operator corrections and scanned pages, and cites the page every value came from.
That is the assumption behind almost every quality system on the market: digitize, template, migrate, and then we can help. It is why so many deployments stall at the point where somebody has to re-key three years of executed records.
Litewave reads what your floor actually produces. Handwritten entries, operator corrections, annotated margins, scanned pages of varying quality, and the same record in three formats across two sites. Every extracted value carries a confidence score and links back to the exact page it came from, and anything below your threshold routes to a human before it feeds a check.
That capability is what makes the rest possible. It is also why the deployment at the world’s largest sulfamethoxazole producer runs on 100 to 150 page paper batch records with no eBPR migration. If your records are already digital, the same checks run against your MES and LIMS data instead. Most sites are somewhere in between, and that is the normal case rather than the exception.
How it works
Your executed record and your master record go in. A review-by-exception file with cited evidence comes out.
- Executed batch record, handwritten, scanned or exported from an eBPR
- Master batch record or MMR
- Equipment, cleaning and line clearance records
- In-process check sheets
- Analytical test records and chromatogram data
- Material issue and reconciliation records
- Exports from LIMS, MES and ERP
- Reads every page into field-level data with confidence scores, handwriting included
- Runs your full check library on every operation in the batch, every time
- Reconciles across documents, not just within one
- Links every extracted value back to its source page
- Pass or fail on every check, exceptions ranked by risk
- Evidence cited inline against each flag
- A drafted review summary for the QA reviewer
- Structured batch data that feeds trending and APQR
What Litewave checks
Configured against your master record, your specifications and the standard your site is inspected against.
Completeness and data integrity
- Every "performed by" and "checked by" field present, signed and dated, with the required separation of roles
- Corrections follow good documentation practice: single line-out, initialled, dated, reason recorded
- No blank fields, no missing pages, page and section sequence intact
- All dates and times sequential and inside the approved processing window
- The executed record matches the effective version of the master record, not a superseded one
Process conformance
- Every recorded parameter, temperature, pressure, pH, RPM, hold time, inside the range in the master record
- Reported operation duration matches the elapsed time between recorded start and end timestamps
- Charged material quantities reconcile between the operations log and the material issue record
- Lot numbers and expiry dates traced to released, in-date material from an approved supplier
- Yield and material balance inside alert and action limits, with variances explained
- Equipment calibration in date, cleaning record and clean hold time within limits
Quality and analytical
- Every specification has a corresponding result, and every result is within limit
- Results in the release record match the analyst raw data, closing the transcription gap
- Reported impurity and assay values reconcile against raw chromatogram data within tolerance
- Residual solvents against ICH Q3C class limits, loss on drying against monograph limits
- OOS and OOT results linked to a closed investigation before disposition
- Environmental and utility monitoring in limit for the manufacturing window
Exception handling
- Every deviation referenced in the record has a corresponding record in the QMS
- Every open item blocking release listed with owner and status
- Change controls affecting the batch identified and their status confirmed
Proven in pharmaceutical production
Measured in production at the world’s largest sulfamethoxazole producer, an FDA compliant and WHO-GMP certified API manufacturer running 100 to 150 page paper batch records, and a mid-size pharmaceutical CMO running 11 automated compliance checks across every operation. Both figures are published in full on the case study pages.
How the World's Largest Sulfamethoxazole Producer Cut Batch Disposition Time by 85%
AI-driven QA acceleration with full audit readiness, cutting batch disposition time by 85% and hands-on review from days to hours across paper-based records, with no infrastructure changes.
Read the case study →How a Mid-Size Pharma CMO Automated 200+ Compliance Checks per Batch
Eleven automated compliance checks across every operation in every batch, every flag backed by source-record evidence, across a constantly changing product mix.
Read the case study →Questions we get asked
Does Litewave release the batch?
No. Litewave performs the review and assembles the evidence. A qualified person on your team makes the disposition decision and signs it in your own system. Litewave records who decided what, on what evidence, and when.
Do we need an electronic batch record system first?
No. Litewave runs on paper and scanned batch records. Our reference deployment runs in continuous production on 100 to 150 page paper batch records, with no eBPR migration and no SOP changes.
How do you handle handwriting?
Field-level extraction with a confidence score on every value, and an engine that adapts to your operators over time. Anything below your confidence threshold routes to a human before it feeds a check, and every extracted value links back to the page it came from so a reviewer can verify it in one click.
Can it run in parallel with our existing review?
Yes, and that is usually how a first deployment starts. Litewave reviews the same batches your team is reviewing manually, and you compare findings for an agreed period before anything changes in your SOP.
What happens to the exceptions it raises?
They arrive ranked by risk with the source evidence attached, so a reviewer is deciding rather than searching. Anything Litewave cannot verify is reported as unverified rather than passed.
Related
Built for validated environments
Your data stays where your auditors expect it
Deploy in your environment
Your cloud, your data center, on the plant floor, or fully air-gapped. The same agents run identically in all four, because the models are open-source and served inside your perimeter, so inference never leaves it.
Your data never trains our models
Documents are never sent to third-party AI providers, and they never train a model that any other customer touches. Where a model does adapt to your site, the handwriting engine learning your operators, that loop is closed inside your deployment and what it learns stays yours.
21 CFR Part 11
Electronic records, electronic signatures and complete audit trails on every action, built to ALCOA+ data integrity principles.
EU GMP Annex 11, and building for what is next
Meets the currently effective Annex 11 expectations for computerised systems. We are building against the draft revised Annex 11 and the draft Annex 22 on AI while both remain in consultation, so nothing here is claimed as compliance with a standard that is not yet in force.
Validated like the computerised system it is
A risk-based validation package written to GAMP 5 second edition and Annex 15 expectations: specifications and traceability matrix, IQ, OQ and PQ protocols with executed evidence, and a supplier assessment pack, so your quality unit can qualify Litewave under its own CSV or CSA procedure.
AI governance you can inspect
Context of use, model versions, performance monitoring and change control are documented against the EMA and FDA guiding principles of good AI practice in drug development, so the model sits inside your quality system rather than beside it.
Evidence on every flag
Every extraction, check and recommendation links back to its source page. Nothing asserted is unverifiable.
Human in the loop by design
Agents propose, qualified people approve. Litewave never dispositions a batch on its own.
See it run on one of your own records
Bring something real. We will show you what Litewave finds in it and what it cites as evidence.