Batch 26-0713 · LC-MS/MS · 214 injections
Every chromatogram, reviewed once.
Cuegent works your batch before anyone opens it: every result classified against your LOD/LOQ presets, every calibration curve gated, every QC recovery checked. Your analyst reviews the exceptions that remain and signs. That is the whole workflow.
This is one injection.
Retention time, response, peak shape. The trace your analyst knows by heart.
This is the batch behind it.
214 injections. Every one needs eyes before results release. Today, three pairs of them.
Cuegent has already worked it.
Detection classified against LOD and LOQ. Calibration curves gated. QC recoveries checked. Seven injections flagged.
▲ 7 flagged for review
You review seven. You sign. The batch releases.
That is one review step instead of three. The capacity was already in your building.
See what it is worth in your labThe review cycle
Your instrument already flags. A person still opens every result.
MultiQuant and SCIEX OS mark ion-ratio failures and outliers. But a flag is not a decision. An analyst still opens each result to judge the flag, reintegrate, and confirm the ones that passed, then a checker repeats the whole pass. That human review of the full batch, done twice, is the work Cuegent removes.
Step 01 · Analyst
First pass
Opens every result, flagged or not. Judges each instrument flag, reintegrates, confirms the passes.
Reviews 100% of results
Step 02 · Checker
Second pass
Repeats the review to catch what the first pass missed. Same screens, same judgment calls.
Reviews 100% of results, again
Step 03 · Reviewer
Final read
Confirms the record is complete and signs the batch for release.
Waits on steps 01 and 02
With Cuegent · One step
Review & sign
Cuegent adjudicates the routine flags against your method and documents the passes. One qualified analyst reviews the exceptions that remain, confirms, and signs.
Reviews only what needs judgment
Cuegent does not replace your instrument's flagging. It does the human adjudication that follows: applying your method's rules to each flag, documenting the routine passes, and surfacing only the results that need an analyst's judgment. Every result still carries a qualified analyst's review and signature.
Agreement
The question is not whether it saves time. It is whether it agrees with you.
Any tool can be fast. The one that matters is whether its adjudication matches the call your own analyst would have made on the same peak. We do not ask you to take a number on faith. We measure it, on your data, against your analysts.
Your analysts set the baseline
During the pilot, your team reviews as they always have. Cuegent reviews the same batches in parallel. Neither sees the other's calls.
Every disagreement is surfaced, not smoothed
The concordance report lists each result where Cuegent's adjudication differed from your analyst's, with the reason and the underlying data. Disagreements are the point: they show you exactly where the method rules need tuning.
You decide whether the agreement rate is good enough
The report is yours. If the rate does not clear your bar, the pilot has told you something true and cost you a batch of parallel review, not a year of licence.
- Results adjudicated in parallelnCount
- Cuegent agrees with analyst%Rate
- Cuegent flagged, analyst passednOver-flaggingReview
- Cuegent passed, analyst flaggednUnder-flaggingReview
- Rule changes made in responsenTuned
Where we stand today: through development with our design partner, no data integrity or result validity issues have been raised against Cuegent's output. That is our partner's assessment, not a measured agreement rate, and we will not present it as one. The rate for your lab comes from your pilot.
Inside the review
The batch arrives worked, not raw.
Every result classified against LOD and LOQ
Each result is classified against your configured presets: Not Detected, below LOQ, or Detected. The classification arrives with the result, not after someone squints at it.
Calibration curves gated, QC recoveries checked
A curve passes only at r² of 0.99 or better, otherwise it is labelled failed to fit. QC recovery is computed against expected and held to your 70 to 130 percent window. Anything outside either gate is surfaced, not buried.
Only exceptions surface
Passing injections are ready and documented. Flags carry the reason they were raised, so review starts at the judgment call, not the click-through.
- Calibration curver² 0.9812 (min 0.99)Below r² gateFlag
- QC recoveryNot evaluatedPass
- Calibration curver² 0.9991 (min 0.99)Pass
- QC recovery62% (tol 70–130%)Below 70%Flag
Audit-ready
Built for audits before the auditor arrives.
Every review decision is timestamped, attributed, and immutable, with the audit trail and export your assessor asks for. Cuegent provides the compliant-capable record layer. Validation in your environment, your SOPs, and configuration remain yours, as they do with any system.
21 CFR Part 11 capable
Electronic records and signatures with a complete, tamper-evident decision history, plus the validation documentation your team needs to qualify it.
ISO 17025 & GLP ready
Role-based access and attributable review actions that map to your existing quality system, not a replacement for it.
One-click exports
The audit trail your assessor asks for, produced in minutes instead of a week of screenshots.
09:41:07 Batch ingested from Analyst 1.7.2 system
09:41:22 Detection classified · 207 of 214 pass system
09:41:22 Calibration & QC recovery gated · 7 flags raised system
10:12:45 Flag reviewed · inj 114 · recovery accepted b.analyst
10:26:03 Flag reviewed · inj 131 · repeat requested b.analyst
10:58:19 Batch signed and released b.analyst
Where we stand
Four numbers, and what stands behind each one.
Three of these are guarantees: they are properties of how Cuegent is built and they do not move. The fourth is a measurement from work with our design partner, and your pilot produces your own.
100%
Results reviewed by a human analyst before release
Guarantee · by design
0
Auto-approvals, ever
Guarantee · by design
2 → 1
Review passes collapsed to one
Guarantee · by design
33–75%
Reduction in review time
Design-partner range · yours is measured in the pilot
Works with your bench
Your instruments, your LIMS, your SOPs.
Cuegent reads the data your systems already produce. No middleware project, no method migration.
Deployment & your data
Where your data sits, and what we do not do with it.
Labs ask this in the first ten minutes, so here it is before you have to ask.
Cuegent runs as a dedicated single-tenant instance per lab on AWS infrastructure that we operate. We do not deploy into your own cloud account or on-premise today, and we will say so rather than imply otherwise. If self-hosting is a hard requirement for you, tell us: it changes our roadmap conversation, not this one.
Common questions
What every lab asks first.
Does Cuegent replace our analysts?
No. Cuegent works the routine review so analysts spend their time on the injections that need judgment. Every result is reviewed and signed by a qualified analyst before release. No auto-approvals, ever.
Is Cuegent compliant with 21 CFR Part 11, ISO 17025, and GLP?
Cuegent provides the compliant-capable record layer: every review decision is timestamped, attributed, and immutable, with role-based access and exportable audit trails, plus validation documentation to support your qualification. Compliance itself is a shared responsibility: validation in your environment, your SOPs, and configuration stay with you, as they do with any regulated system.
Which instruments and LIMS does Cuegent connect to?
SCIEX Analyst / MultiQuant (LC-MS/MS) and Agilent MassHunter (GC-MS) are supported today. Shimadzu LabSolutions is in validation. QBench is the supported LIMS. Thermo Fisher and Waters are on the roadmap: if that is your stack, the calculator below will tell you and we can notify you when support lands.
What happens when Cuegent gets it wrong?
It gets surfaced, not buried. Cuegent proposes; it never silently applies a reintegration or approves a result. Every decision it makes is visible with the reason attached, so a wrong call shows up as something your analyst overrules in the normal course of review. During the pilot, those overrules are counted and reported back to you as the disagreement rate, and they drive the rule tuning. A tool that hides its errors cannot be checked. This one is built to be checked.
Do you train your models on our data?
No. Your data is not used to train models. If that ever changes, we will ask for your written permission first, and declining changes nothing about your service. Whatever the site says here will match what the contract says: if you find a gap between the two, that is our error and we want to hear about it.
Can we run Cuegent on our own servers or in our own cloud?
Not today. Cuegent runs as a dedicated single-tenant instance per lab on AWS infrastructure we operate. Deployment into your own VPC or on-premise is not something we support at this stage. If it is a hard requirement, say so early and we will tell you honestly whether we are a fit.
How long does onboarding take?
Cuegent maps to your existing QC rules and instrument setup during onboarding. There is no method migration and no new process for the bench: it reads the data your systems already produce. A pilot typically runs one to three months, and it begins with a file check before any commitment.
How a pilot works
Four steps, and the first one costs you nothing but a file.
We do not start with a contract or a deployment. We start by proving we can read your data, because if we cannot, nothing else matters.
01
Send us sample files
Raw acquisition and result files from your instrument. We verify our ingestion pipeline reads your methods, your transitions, and your calibration and QC structure correctly, and we tell you plainly if it does not.
Before any commitment
02
Onboard and configure
We stand up your instance and configure Cuegent against your method rules, tolerances, and SOPs. No method migration, no change at the bench.
Your rules, not ours
03
Run it against real batches
Your analysts review as they always have. Cuegent runs in parallel on the same batches. We tune alignment as the disagreements teach us where the rules need work.
Typically one to three months
04
Read the KPIs and decide
You end the pilot holding a concordance report and your own review-time numbers, measured on your own batches. Then you decide.
Evidence, then commitment
Pilot commercial terms are set per lab. Ask us and we will give you a straight answer on the call rather than a number on a page that will not fit your situation.
Pilot fit & ROI
What a single review step is worth in your lab.
The first four questions score your pilot fit against what Cuegent supports today. The rest model your review workload the way it is actually structured: chromatogram review across every pass, plus the curve and QC prep that happens once per batch. Every assumption behind the numbers is listed at the bottom.
Fit is scored against the instruments and CDS platforms Cuegent supports today.
3,917
Current review hrs / yr
2,434
With Cuegent
Where the time is saved
1,482
Hours returned / year
148
Hours back / analyst / year
$77,820
Labour saving / year
0.74 FTE
Capacity recovered
+3,312
Added sample capacity / year
$149,018
Revenue from freed capacity / yr
You stay in control. Cuegent automates data review and review prep. Exceptions and the final decision stay with your analysts.
Not a fit today. Leave your stack details and we will contact you when support lands.
Get notified