Lab Agent OS · for regulated labs

The paperwork runs itself. You sign what matters.

A crew of compliance-grade AI agents for deviations, CAPA, SOPs, training and audit prep — every step traced to its source, every decision signed by a person.

Most quality work is assembling evidence that already exists in five systems, in the right format, by a deadline. Lab Agent OS drafts the deviation, scopes the impact, proposes the CAPA and routes it for signature — inside the LIMS and QMS you already run.

shield

21 CFR Part 11

shield

EU GMP Annex 11

shield

GAMP 5 · CSA

shield

SOC 2 Type II

DEV-2026-0412 · Deviation Agent● run 7f3a·c2
09:14:02SENSEHPLC-07 column oven variance ±1.8 °C over 40 min (limit ±1.0)source: Trucision instrument telemetry
09:14:05PLANMeets deviation criteria per SOP-QA-014 §5.2 → open record, scope impactretrieved: SOP-QA-014 Rev 6, effective 2026-03-02
09:14:09ACTLIMS query: 3 sequences on HPLC-07 since 08:31 → lots 24-1187, 24-1188tool: lims.search_runs · read-only
09:14:16ACTDrafted deviation; attached 72 h instrument health history
09:14:21VERIFYChecker agent: 6/6 citations resolve · 1 gap (system suitability result missing) → revisedsecond model, independent context
09:14:30RISKPreliminary: Major · severity 3 × occurrence 2 (ICH Q9 matrix)
09:14:31HOLDLots flagged “pending QA review” in LIMSreversible action · autonomy L2
09:14:32ROUTESent to QA reviewer M. Okafor for e-signature
○ Awaiting signature · meaning: ReviewOpen for review

Illustrative run: 30 seconds from instrument signal to a cited deviation draft. The QA decision stays human.

The work between the science

Your QA team didn’t train for years to chase signatures.

Most quality work is assembling evidence that already exists in five systems. Agents are good at exactly that — and it’s the work nobody on your team will miss.

shield icon

Deviations & investigations

Today: the analyst writes it up from memory, then QA rewrites it.

bullet point

Drafts the deviation from the instrument signal, LIMS runs and the governing SOP, with every claim cited

bullet point

Builds a 5-Whys / Ishikawa from telemetry, maintenance and training records

bullet point

Proposes CAPA with measurable effectiveness checks and tracks every due date

bullet point

Autonomy L1: drafts for a person to edit and submit

clock icon

SOPs & change control

Today: doc control redlines by hand after every change.

bullet point

Finds every SOP, form and training module a change touches

bullet point

Produces tracked redlines and an impact summary reviewers actually read

bullet point

Flags conflicting SOPs and documents overdue for periodic review

bullet point

Autonomy L1: drafts for a person to edit and submit

server icon

Training & calibration

Today: managers cross-check the training matrix; planners juggle fixed intervals.

bullet point

Assigns read-and-understand on the SOP effective date and chases overdue items

bullet point

Blocks unqualified analysts from running a method in LIMS

bullet point

Schedules calibration from predicted drift and books the service window

bullet point

Autonomy L2: reversible, logged actions

thumbs-up icon

Audit readiness

Today: two weeks of binders before every inspection.

bullet point

Runs a mock inspection every night against your SOPs and regulations

bullet point

Keeps a live evidence room that answers “show me” requests with source records

bullet point

Drafts responses to auditor requests during the inspection

bullet point

Record approval stays human: agents never apply an e-signature

How an agent works

Not a chatbot. A closed loop with a checker and a signature at the end.

Every agent runs the same governed loop: grounded in your SOPs, acting only through connectors you’ve scoped, checked by a second model, and stopping where a regulated decision is made.

Sense & reason — grounded in your procedures

Triggered by an instrument alarm, an OOS result, an SOP change or an overdue CAPA. The agent pulls the effective revision of every SOP, spec and prior record that governs the event, then plans against them.

Act — inside the systems you already run

Queries and updates LIMS, ELN, QMS and LMS through scoped, least-privilege connectors. No copies, no parallel system of record, nothing to migrate.

Verify — checked before anyone sees it

An independent checker model resolves every citation, tests the draft against acceptance criteria and sends weak work back. Low confidence goes to a person, not forward.

Sign — a qualified person decides

The record arrives complete, cited and risk-ranked. Approval is a Part 11 e-signature by an authenticated human. Reviewer edits feed the agent’s evaluation set — never silently into the live model.

The crew

One agent per quality process. One operating system underneath.

Agents hand work to each other — a deviation spawns a CAPA, a CAPA triggers an SOP change, the SOP change triggers training — so one event is followed all the way to closure.

Deviation agent

Turns alarms, OOS/OOT results and analyst notes into a cited deviation draft.

bullet point

Reads: Trucision instrument alerts, LIMS results and run history, analyst notes

bullet point

Produces: deviation record, affected lots and samples, ICH Q9 preliminary risk

bullet point

L1 draft · 21 CFR 211.192 · ISO 17025 §7.10

CAPA agent

Runs the investigation workbench and chases CAPA to closure.

bullet point

Reads: open deviations, instrument health history, training and maintenance logs

bullet point

Produces: root-cause analysis with evidence, CAPA plan and owners, effectiveness checks

bullet point

L1 draft · ICH Q10 · ISO 13485 §8.5.2 · ISO 17025 §8.7

Change control agent

Maps every document, validation and training record a change touches.

bullet point

Reads: change requests, equipment and method master data, validation inventory

bullet point

Produces: impact assessment, revalidation checklist, linked SOP redline tasks

bullet point

L1 draft · EU GMP Annex 15 · ICH Q10

SOP & documents agent

Keeps controlled documents coherent and current.

bullet point

Reads: controlled document library, approved change requests, regulatory updates

bullet point

Produces: tracked redlines, conflict and gap report, periodic review queue

bullet point

L1 draft · 21 CFR 211.100 · ISO 13485 §4.2.4

Training agent

Assigns, tests and enforces analyst qualification.

bullet point

Reads: role and training matrix, SOP effective dates, LIMS method assignments

bullet point

Produces: assignments and reminders, competency checks generated from the SOP, qualification flags

bullet point

L2 reversible · 21 CFR 211.25 · ISO 17025 §6.2

Calibration & maintenance agent

Condition-based schedules, built on Trucision Predictive Maintenance.

bullet point

Reads: predicted drift and health scores, sample queue, service contracts

bullet point

Produces: condition-based schedule, vendor bookings, equipment log entries

bullet point

L2 reversible · 21 CFR 211.68 · ISO 17025 §6.4

Audit readiness agent

A mock inspection every night, and a live evidence room.

bullet point

Reads: all quality records, regulations and sponsor SOPs, past findings

bullet point

Produces: daily readiness score by area, evidence packets, draft responses

bullet point

L1 draft · ISO 17025 §8.8 · FDA inspection readiness

Supplier & materials agent

Checks CoAs against spec and tracks supplier qualification.

bullet point

Reads: CoAs and incoming inspection, supplier master, material specs

bullet point

Produces: CoA check results, requalification calendar, draft SCARs

bullet point

L1 draft · 21 CFR 211.84 · ISO 13485 §7.4

Governed by design

Validate the agent like you validate an instrument.

Bounded autonomy you set per workflow, a validation package in line with GAMP 5 and Computer Software Assurance, and a record an inspector can follow step by step. In critical GMP decisions, AI assists and people decide.

Bounded autonomy

Start every agent in shadow mode. Promote it one level at a time as its record earns it.

bullet point

L0 Shadow: runs silently beside your team; output is compared, never shown

bullet point

L1 Draft · L2 Act reversibly · L3 Close routine items within pre-approved rules

bullet point

Never: sign, approve or release. Hard-coded, not a setting

Validation package

An agent has a configuration, an operating range and a performance record. We deliver the evidence with it.

bullet point

IQ: model, prompts, tools and scopes hashed and locked under change control

bullet point

OQ: replays hundreds of your closed deviations and CAPAs against agreed criteria

bullet point

PQ: 30 days in shadow on live work. Drift past a threshold drops the agent a level

Explainable to an inspector

“How did you conclude that?” — open the run. Every step is kept as an ALCOA+ record.

bullet point

Agent ID, model version and approving user on every record

bullet point

Every claim links to the exact SOP revision, LIMS result or telemetry window

bullet point

Append-only audit trail; reviewer edits shown as diffs against the draft

Your stack, your data

No new system of record. Records stay in your validated QMS and LIMS.

bullet point

Connects to QMS, LIMS/ELN, document control, LMS, ERP and Trucision telemetry

bullet point

Private deployment; your records are never used to train foundation models

bullet point

Pause any agent, site or workflow instantly; work in flight returns to a human queue

30-day pilot

Bring us one deviation. We’ll show you the agent that would have written it.

Pick one workflow. In 30 days you’ll have a shadow-mode agent working on your real records, and the validation evidence to decide whether it goes live.

Week 1 — Scope

Choose the workflow, map SOPs and connectors, and agree acceptance criteria with QA.

Week 2 — Replay (OQ)

Run the agent on your closed records and review the scored results together.

Weeks 3–4 — Shadow (PQ)

Run at L0 on live work, trending agreement and reviewer edits every day.

Day 30 — Decide

You get a validation summary and a go/no-go for promotion to L1.