Let’s talk
All case studies

AI research system

TradeLab

A research workbench that connects hypotheses, source evidence, critique and simulated decisions.

My contribution
System design · agent orchestration · evidence boundaries
Timeframe
Iterative research platform
Status
Ongoing
THE WORK IN ONE LINE

Research, critique and source evidence stay connected in a reviewable workflow. Shadow-only research prototype.

PythonFastAPIReactPostgreSQLTemporal

01 / The problem

More agents should not mean more unverified claims.

Multi-agent research can produce volume without trust. The design problem was to make delegation legible, preserve provenance, and prevent unsupported conclusions from silently becoming recommendations.

02 / What I built

The workflow assigns distinct responsibilities to eight roles, records evidence as structured artifacts, and makes review states visible. The replay is deterministic and represents a prerecorded synthetic session.

Prerecorded synthetic session

Inspect the research system

Fictional subject. No account or brokerage connection.

Research graph

Eight bounded roles

Read-only
Lead
analyst
Planner

Breaks the question into bounded research assignments.

Produces: Research plan

Evidence ledger

Reviewable provenance

Synthetic evidence ledger
SourceTypeOwnerStatus
EV-101Synthetic quarterly filingDocumentData CollectorVerified
EV-102Fictional industry outlookReportDomain ExpertVerified
EV-103Generated competitor tableDatasetFact CheckerVerified
EV-104Counter-thesis memoAnalysisSkepticUnder review
EV-105Assumption registerLedgerArchivistVerified

Workflow timeline

Planning

Planner

Question decomposed and source boundary approved.

04 / Result & scope

What changed. What comes next.

Sanitized architecture and workflow artifacts from the TradeLab development workspace.

Eight bounded roles

Separated planning, evidence collection, challenge, synthesis, and presentation responsibilities.

Evidence-first output

Designed a ledger that keeps research artifacts attributable and reviewable.

Fail-closed boundary

Kept the public and research experience read-only with no execution capability.

01

Make the hypothesis explicit

A research objective becomes a tracked hypothesis with evidence and review states. Structured contracts keep the UI and workflow in agreement.

02

Keep the source with the claim

An evidence ledger records where a finding came from and whether it survived critique. Stale or contradictory material remains visible rather than disappearing into a confident summary.

03

Research scope

TradeLab v2 is shadow-only with no broker-write capability. Product acceptance remains open. The public replay demonstrates a fictional session, not investment performance.

Continue exploring

All case studies