Bring the whole trial into the review.

Connect trial data, safety records and study documents. Ask neil to find what is missing, prepare the review and prepare the follow-up in Seal.

Book a demo
Illustration of a seal beside a sample transport case and study records.
You:

What is still stopping database lock for CL-042? Prepare Friday’s review from the EDC, safety reconciliation and coding records.

neil

There are 18 open queries, but the count alone misses the problem. Eleven are more than 14 days old, concentrated at two sites. I also found an unmatched safety record and an unfinished coding review.

Open queries by age11 of 18 over 14 days

Each dot is an open query. Red marks the study’s follow-up threshold—not clinical severity.

Start with sites 101 and 309. Closing their queries will not, by itself, complete the lock checklist.

About this example

A prepared conversation using fictional records, not a live query. Source buttons open the example evidence. The draft can be read and downloaded; no records are changed, shared or approved.

Your trial can stay in its existing systems.

Ask one question across EDC, CTMS, safety and essential documents. neil can examine permitted records through scoped connections or supplied exports, then bring the relevant versions into a reviewable answer.

Keep the EDC as the source of case-report data and the safety system as the source of safety cases. A summary should retain study, site and record identifiers, source time and any missing coverage. Reconciliation is visible work, not a promise that connecting systems makes discrepancies disappear.

Carry the finding into the follow-up.

Configure the study-review workflow in Seal: a finding, its source records, the responsible role and the evidence needed to resolve it. A reviewer can follow the answer back to the records and forward to the outstanding work.

Use the same approach for protocol deviations, amendment rollout and inspection requests. Keep a proposed query distinct from one actually issued; a site response distinct from reviewer acceptance; and a prepared review package distinct from approval. Sponsor, CRO and site responsibilities remain explicit.

Start with one study review.

Bring the checklist and a bounded set of records. Agree the identities, source systems, access and output before connecting the wider trial.

Test a stale extract, an answered-but-open query, an unmatched identifier and a restricted source. The demonstration should show what neil found, what it could not inspect, and which statements a reviewer can reproduce from the evidence. Study teams retain control of data changes and decisions.

Explore the configured capabilities

Study review

Bring source-time-aligned EDC, safety and coding records into one review. Keep unresolved discrepancies and missing source coverage visible.

Electronic data capture

Configure forms, edit checks and query workflows for the study, or connect selected records from the existing EDC.

Explore this blueprint

Trial operations

Connect site plans, monitoring findings, delegated work and sponsor review to the study.

Explore this blueprint

Essential documents

Link the required document versions to study events, expected-record checks and review work.

Explore this blueprint

Protocol deviations

Retain the applicable protocol version, affected visit or site, assessment and follow-up. A possible discrepancy is not automatically a confirmed deviation.

AI review assistance

Ask neil to draft source-backed review briefs and query proposals. Reviewers decide what to issue, accept or change.

Explore the connected records

Questions worth asking.

No. Start with a scoped connection or supplied exports from your existing EDC. Agree the identifiers, fields, permissions and update frequency. Seal can provide the connected review and follow-up workflow while the EDC remains authoritative for case-report data.
A review brief with source references, unresolved evidence gaps and proposed follow-up work. It can also draft queries or a configured workflow. Those proposals remain distinct from issued queries, changed study data and authorised decisions.
Yes. Scope the forms, edit checks, roles and review states for the intended study. Verify the configured process, integrations and required exports before operational use; do not assume that one generic form set fits every trial.
Keep safety-case ownership and reporting responsibilities explicit. Connect the permitted records needed for reconciliation and oversight. An AI summary is not a medical assessment, a completed reconciliation or an expedited safety submission.
Define access by study, role and permitted source scope, including any blinding restrictions. Test both allowed and denied access before use. The review should identify incomplete coverage rather than implying that unavailable records were checked.
Use one study-review checklist and a bounded set of records. Check the source links, identifiers, stale extracts and discrepancy handling. Agree the intended use, validation approach, training and accountable reviewers before expanding.

Bring the study that needs a clearer review.

Tell us about the output your team needs. We’ll show you the workflow and discuss the records and connections it would use.

Book a demo
Example source record