Private by default
Programme documents are isolated from the public synthetic run and direct browser table access.
Starting-cell decision intelligence
Why this cell line?
Riyaan brings candidate evidence, rights, review status, and decision history into one traceable workspace for allogeneic iPSC programmes.
For cell-therapy and CMC leads choosing allogeneic iPSC starting lines. Named experts approve every consequential call; Riyaan does not pick the scientifically best line.
Public selection run · fully synthetic
One question carried across five days.
Selection question approved
Programme scope and review boundary fixed.
New source arrived
Synthetic source bundle revision 3 registered.
Evidence extraction drafted
Agent task completed; claims remain unreviewed.
Named review resolved
Source-linked rights evidence approved.
Candidate excluded
Reporter modification confirmed for DEMO-iPSC-C03.
Shortlist brief approved
Revision 2 retained with its evidence snapshot.
Line selection takes weeks
The decision does not happen in one meeting.
Source records, reviewers, proof results, and deadlines arrive asynchronously. The selection question stays open while the evidence underneath it keeps moving.
- Sources change
- A registry record is revised, a report arrives, or a source conflict appears.
- Owners change
- A new reviewer takes over while evidence and rationale remain distributed.
- Conclusions change
- A proof outcome advances, holds, or excludes a candidate days later.
What Riyaan owns
The work between the question and the review.
Riyaan keeps an approved mission moving across days or weeks, then stops at the boundary where expert judgment is required.
- 01
Gather the evidence
Register approved sources, snapshots, versions, and retrieval dates.
- 02
Maintain the gap queue
Keep missing, conflicting, unreviewed, and expired evidence visible.
- 03
Monitor approved sources
Resume when a source changes, a deadline arrives, or a result is recorded.
- 04
Rank the next proof
Use deterministic programme rules to surface the proof most likely to change the call.
- 05
Assemble the brief
Produce an immutable, source-linked shortlist revision for expert approval.
Human review boundary
The agent moves the work. Experts own the call.
Riyaan may
Prepare and propose
- Gather and version approved sources
- Draft source-cited extractions as unreviewed
- Apply deterministic readiness and blocker rules
- Propose the next proof and assemble a brief
Named experts
Review and approve
- Evidence promotion into the canonical record
- Proof plans, acceptance criteria, cost, and timing
- Candidate advance, hold, or exclusion
- The final shortlist—or “no line is ready”
Riyaan does not autonomously choose the scientifically “best” line or make a clinical recommendation.
Finished artifact
A shortlist brief the team can review, approve, and retain.
Each revision binds the candidate dispositions to the evidence snapshot, protocol version, unresolved blockers, and named approval behind it.
- Candidates and exclusions
- Source-linked rationale
- Unresolved risks
- The next proof worth funding
Shortlist brief · revision 2
Synthetic starting-cell selection
DEMO-iPSC-A01AdvanceIdentity, quality, and rights evidence approved.
DEMO-iPSC-B02HoldCommercial-use rights remain unresolved.
DEMO-iPSC-C03ExcludeReporter modification confirmed.
Unresolved blocker
DEMO-iPSC-B02 rights packet
Next proof worth funding
Resolve the DEMO-iPSC-B02 rights packet
Fully synthetic example · not a scientific or clinical recommendation
Private-pilot trust model
Private programme work stays separate from the public run.
Private ingestion and managed-model processing remain disabled until the pilot's authentication, retention, storage, deletion, audit, and backup controls are approved.
Scoped access
Passwordless identity, organization membership, programme scope, and named reviewer roles.
Fail-closed processing
Quarantine, source allowlists, citation validation, and explicit model-retention approval gates.
Auditable work
Signals, tasks, approvals, artifact revisions, and safe metadata stay in the mission record.
Want to inspect the boundary before sharing data?
Founding pilot
Bring one live starting-cell selection question.
We will map the sources, review gates, open proofs, and finished brief your programme needs—without claiming Riyaan can replace the experts responsible for the decision.