QC that moves the batch forward.

A pharmaceutical QC laboratory controls more than samples and results. It controls the relationship between the material, specification, method, analyst, instrument, raw data, calculation and review behind each result, and the manufacturing decision that follows.

Seal runs that relationship as one chain. Incoming, in-process, environmental, stability and finished-product samples enter with their source and requirements attached. Instrument data keeps its context, results are checked against the applicable specification, and approved decisions return to the batch without transcription.

1One result needs its whole context.

The value 98.7% cannot support a release decision by itself. A reviewer needs the sample and preparation that produced it, the method version, instrument and acquisition, the standard and its potency factor, the calculation, the specification version and who performed and reviewed the work. They also need to know whether anything was reprocessed or changed.

FDA’s data-integrity guidance expects complete, consistent and accurate data, with the metadata needed to reconstruct the activity.¹ In Seal the laboratory record is not a typed summary of the instrument record. It references the source, preserves each transformation and exposes the review trail from acquisition to disposition.

1.1Why teams choose Seal for QC

Samples, methods, results and batches share an ontology across QC, manufacturing and quality in Seal. Your team can ask Neil to build a method workflow, compare results or investigate an exception with its original context. Findings can become revised calculations, checks or review steps for testing and approval. Earlier results retain the method and specification used, while new work follows the released version. Laboratory improvements carry through to the operation that depends on them.

2Samples arrive ready to route.

Samples originate in different operations but enter one controlled intake:

• receiving creates incoming-material samples from the approved sampling plan;
• manufacturing steps create in-process and release samples with batch priority;
• environmental schedules create viable, surface, personnel and utility samples;
• stability protocols create pulls and test assignments at defined timepoints; and
• investigations create targeted samples with a stated rationale and disposition rules.

Each sample carries its material and batch identity, source, storage conditions, test panel, specification, due date and chain of custody. Accessioning confirms what physically arrived instead of asking an analyst to reconstruct why it exists. Work queues can reflect manufacturing need, hold times, stability windows, instrument availability and analyst qualification, so priority is recorded state rather than a coloured cell in a planning spreadsheet.

3Specifications and methods are executable controls.

A specification should determine behaviour, not sit beside the result as a PDF. Seal holds tests, acceptance criteria, calculations, units, rounding and effective versions as approved data. When a sample is created, the applicable specification resolves from the material, stage, market, method and effective date. The analyst sees the tests to perform, and a result outside limits is flagged for assessment rather than passing silently through review.

Historical samples keep the specification applied when the work was performed. A proposed change identifies the affected products, methods, worksheets, stability protocols and future samples before approval.²

An approved method becomes a controlled execution pattern: preparations, standards, system suitability, sequence rules, calculations and signatures in the order the work requires. Configured checks can confirm analyst qualification and instrument state before work begins. Standards keep their lot, potency, expiry and storage; calculations use approved formulas and record their inputs; failed suitability follows the defined response. The system does not replace scientific judgement. It shows the reviewer where judgement was applied and what evidence supported it.

4The instrument output remains the original.

Instruments range from a balance to a chromatography data system or a file-producing analyser, and the principle is the same for each: people should not retype a machine-generated value. Seal captures data through an appropriate interface, file watcher, API or controlled import. The record keeps the instrument, acquisition time, source file, parser or mapping version and the extracted values. Where the authoritative raw data stays in a specialist data system, Seal keeps a secure reference to it for review rather than treating a copied number as the original record.

Manual entry remains for observations that are genuinely manual, with unit and range checks, attribution at the time of entry, a reason for any change, and review.

5Investigations open with the evidence attached.

An out-of-specification result opens an investigation with the analytical context already linked: sample, method version, preparation, instrument, sequence, standard, analyst, calculations, audit trail, neighbouring results and specification.

The laboratory runs its approved initial assessment, hypothesis testing, retest or resample controls and any manufacturing investigation while original and additional data remain distinct.³ Every added result states why it exists and how it affects the conclusion. Out-of-trend and atypical results use the same connected history, so reviewers can investigate drift while results are still within specification.

6Stability and environmental monitoring run from their programmes.

A stability study connects the protocol, batch, packaging configuration, storage condition, chambers, pulls, tests and specifications. Seal generates the work from the approved protocol rather than from a calendar kept beside the LIMS.⁴ Chamber excursions identify the samples and exposure windows affected, missed or partial pulls remain visible, and results trend across timepoints and batches using the same methods and specifications as the rest of QC.

Environmental monitoring follows the same pattern. The monitoring programme governs locations, frequencies, limits and trending; the laboratory governs receipt, media, incubation, enumeration, identification and review. The schedule creates laboratory samples, and microbiology returns reviewed results to the location and time window they describe, where organism identities and recurring recoveries inform facility trending and deviation assessment.

7Render the CoA from approved data.

A Certificate of Analysis presents approved data: product and lot identity, manufacturing or retest dates, tests, specifications, reported results, status and authorised approval. Seal generates it from the release dataset. Customer or market templates control presentation without creating a second copy of the result, and an amendment creates a new controlled version with its reason.

The disposition that makes the lot available can make its approved CoA available too, without a separate document queue after release.

8Stand up the lab around one complete sample journey.

A QC implementation should prove the full chain, not a list of screens. Select representative incoming, in-process, finished-product, environmental and stability samples, and configure their specifications, methods, instruments, calculations, review and downstream decisions.

Exercise failures deliberately: missing chain of custody, an expired standard, overdue calibration, failed suitability, an OOS result, a corrected entry, a chamber excursion and an amended CoA. The laboratory is ready when normal and exception work both produce complete, reviewable records without side spreadsheets.