Cross-system contradictions
Gaps that exist between systems rather than inside any one of them — where no single platform can see the problem.
WHAT NOTHING ELSE SURFACES CROSS-SYSTEM EVIDENCE FOR ICH E6(R3)
You declared
what's critical.
Can you show the controls operated?
E6(R3) asks you to name your critical-to-quality factors and a control for each. The evidence that a control actually operated almost never sits in one system — it is a relationship between records held by different vendors. Vigil reads the exports you already produce and reports where that evidence cannot be found.
✧Sponsor-side Quality and Clinical Operations · Design-partner conversations open
VIGIL INTELLIGENCECONCEPT VIEW•••
DECLARED CONTROL SWEEP
Six controls checked
Illustrative concept · Not a live product screen
SWEEP RESULT
5 not evidenced →
Each finding cites the records behind it31SUBJECTS
EligibilityRandomised too early
Control 02Not evidenced
Control 03Operating
Two dates, two vendors
01Randomisation preceded the confirming read. Neither system can see the other.
Every finding cites its records
SPONSOR OVERSIGHT EVIDENCEDECLARED CONTROLS · CROSS-SYSTEM CHECKS · CITED RECORDSSCROLL TO EXPLORE ↓
01 / THE CHALLENGE
Every system reports on its own work.
Your eTMF is complete on its own terms. Your CTMS logged no deviation, because from inside the CTMS nothing deviated. A monitoring platform marks its own signal closed when somebody closes it there. Each answer is true, and none of them is independent.
So the question is the next one: what tells you, today, that the controls you declared are still operating?
IN FORCE SINCE 1 OCTOBER 2026ICH E6(R3)
Health Canada now enforces ICH E6(R3) in full. The FDA has adopted E6(R3) as guidance without a compliance date, so the obligation to evidence sponsor oversight is sharper in Canada today than in the United States. Vigil is being developed for sponsor-side oversight in both markets. European (EMA) and broader EMEA contexts are being considered for a future phase—not offered today.
01
The evidence spans systems
Proof that a control operated sits across the TMF, the CTMS, safety, the QMS and vendor records at once.
02
No system audits itself
A platform marking its own record complete is reporting on its own work, however good the platform.
03
Found late, or found for you
An inspector reads across systems the way nobody inside the organisation ever has.
02 / HOW IT WORKS
From declared controls
to evidence they operated.
You export the files you already produce. Vigil reconciles them against each other and against the controls your risk assessment declares, and reports what it cannot find — with the records cited.
STAGE 01 — FILE-BASED INGESTIONVIGIL / CONCEPT
The exports your team already runs.
Standard CSV from your TMF, CTMS and the other systems that hold oversight evidence. Vigil holds no credentials to your systems, opens no connection to them, and needs no firewall change on your side.
TMFCTMSQMS
VIGILINTELLIGENCE LAYER
FindingsCited records
●This illustration communicates intended outcomes, not Vigil's technical implementation.
WHAT COMES BACKFindings you can
Findings you can
put in front of an inspector.
Not an impression from a dashboard. A record naming the rule, the threshold, the regulatory basis and the underlying documents.
Checks against your own controls
Your risk assessment names what is critical. Vigil checks whether the control you declared for it is the one actually operating.
CRITICAL-TO-QUALITYEvidence chains on demand
Inspector-style questions answered with the supporting records assembled and the gaps named rather than hidden.
DEFENSIBLE, NOT ASSERTED 03 / WHAT THIS LOOKS LIKE
Every system complete.
The gap between them.
An illustrative example. A declared control requires the central laboratory to confirm the screening read before a subject is randomised. For 31 of 240 subjects, the randomisation timestamp precedes the confirming report date.
WHAT EACH SYSTEM SAYSAll correct.
▤
All correct.
All complete.
Randomisation ran every subject it was asked to and cannot see an external laboratory. The laboratory returned every report inside target and does not know when anyone was randomised. The CTMS logged no deviation, because nothing deviated inside it.
NOTHING TO DETECT, SEPARATELY→Comparedone record against another
WHAT THE COMPARISON SHOWSA control that
◎
A control that
never operated.
Two dates held by two vendors, in the wrong order, measured against a requirement written in a third system. Not a missing document — a declared control on a declared critical factor, with no evidence it was ever in force.
31 OF 240 SUBJECTSThese values are individually unremarkable. The affected subjects sit inside the protocol window, their readings are ordinary, and their sites enrol at normal rates. Nothing here is a statistical outlier, which is why statistical monitoring does not reach it.
An anomaly is a value that is unlikely.
A contradiction is a claim that is unsupported.
Finding one needs variance. Finding the other needs a join.
A contradiction is a claim that is unsupported.
Finding one needs variance. Finding the other needs a join.
Conceptual illustration drawn from a synthetic study. Not a real study result, and not a claim of validated inspection performance.
01 / FOUND EARLYA documented decision
An impact assessment and a rationale, on your own initiative, with dates proving when.
02 / FOUND LATEA disclosure under pressure
The same finding, with the remediation window gone and the elapsed time on the record.
03 / FOUND FOR YOUA question about the system
An inspector asks whether the control operated. The subject is no longer the finding.
A COMPLEMENTARY INTELLIGENCE LAYER
Not another system
to replace your systems.
Your TMF stays your TMF. Nothing migrates, nothing is replaced, and you can extract everything you put in at any time. Vigil holds no credentials to your systems and opens no connection to them, so there is no integration project and no security review of a live link.
Start with one study →✦
VIGILSponsor oversight perspective
INTELLIGENCEeTMF & CTMSDocuments and execution
Safety & QMSEvents, deviations, CAPAs
Vendor recordsDelegated oversight
04 / BUILT FOR BETTER DECISIONS
Clarity for the people
accountable for quality.
Built for mid-market sponsors running studies across systems they do not all control—starting with sponsor Quality and Clinical Operations.
Quality and QA leads
Check whether the controls your own risk assessment declares are the ones actually operating, before someone else checks for you.
CRITICAL-TO-QUALITYClinical Operations
See where a documentation chain stops halfway, and which sites a monitoring plan no longer fits after an amendment.
OPERATIONAL CONTROLSponsors running CRO-led studies
Evidence the oversight of delegated work when the systems holding the proof belong to somebody else.
DELEGATED ACCOUNTABILITYRegulatory Affairs
Assemble the evidence behind an inspector-style question in advance, with its gaps named rather than discovered.
INSPECTION READINESS RESPONSIBLE BY DESIGN
Intelligence should make judgment stronger.
01
Human oversight stays central
Vigil aims to inform expert review; it does not automate regulatory or clinical judgment.
02
Every finding is explainable
A stated rule, a stated threshold, a citable regulatory basis and the records behind it. No machine learning sits in the detection path, because a finding you cannot defend to an inspector is worse than none.
03
Vigil never cites its own output as evidence
Every evidence reference resolves to a record from your systems, never to another Vigil finding. Independent assessment is the claim; it would not survive assessing our own work. Real-world performance must be demonstrated before claims about impact are made.
05 / DESIGN-PARTNER PATHWAY
One study.
Two exports.
One session.
We run the analysis against your own data and walk you through every finding. If nothing it surfaces is worth acting on, you have lost an afternoon and we have learned something.
01
SCOPEPick one study
Agree the critical-to-quality factors to check, and the data safeguards, before anything moves.
02
RUNTwo exports
A standard TMF export and a standard CTMS export. No system access, no credentials, no IT involvement.
03
REVIEWWalk the findings
Go through each finding and its underlying records with your Quality and Clinical Operations leads.
One afternoonIndicative scope for a first look; any wider evaluation is scoped jointly
THE IMPORTANT QUESTIONS
Clear answers.
No inflated promises.
Is Vigil a replacement for eTMF, CTMS, QMS, or RBQM?+
No. Your systems of record stay your systems of record. Vigil reads copies of what they export and reports on the relationships between them. Nothing migrates, and you can extract your data at any time.
We already have central statistical monitoring. Isn't this the same thing?+
No, and the two are complementary rather than alternative. Statistical monitoring looks for values improbable relative to other values — enrolment outliers, digit preference, implausible variance — and nothing substitutes for it where a distribution exists. The gaps Vigil looks for have no distribution: a date in one vendor's system in the wrong order relative to a date in another's is not an outlier in any population. An anomaly is a value that is unlikely; a contradiction is a claim that is unsupported.
Does Vigil predict which studies regulators will inspect?+
No. Regulatory inspection selection is driven by factors beyond a sponsor's internal systems. Vigil aims to support better-informed oversight and inspection-readiness planning, not forecast regulator behavior or guarantee outcomes.
Is the platform already commercially validated?+
No. Vigil is at the design-partner stage. A working detection engine runs against a synthetic study; real-world commercial performance remains to be established through responsible evaluation. The illustrative findings shown on this page are engineered to exercise specific detection rules — they demonstrate what the engine looks for, not how often such gaps occur.
Does Vigil need access to our systems, or our patient data?+
Neither. Ingestion is file-based: you upload standard exports. Vigil holds no credentials to your systems, opens no connection to them, and requires no network or firewall change on your side. No personal health information is ingested at any point — subject identifiers are study-internal only.
Which regulatory markets is Vigil considering?+
Our initial focus is sponsor-side clinical oversight in the United States and Canada, informed by ICH GCP principles. European regulatory and broader EMEA considerations may inform a future phase; those services are not currently offered.
How would an initial evaluation work?+
We begin with a bounded business question, agree on the evaluation scope and data safeguards, and review usefulness with your Quality and Clinical Operations leads.
THE NEXT CONVERSATION
Most of the evidence
already exists.
It just is not in one place. Send one study's exports and we will show you what is sitting between your systems.
Start a design-partner conversation →sandy@vigilclinical.comVigil is in design-partner development. Product performance and security arrangements require evaluation; no regulatory outcome is guaranteed. Do not send clinical trial records by ordinary email.