⚠ SYNTHETIC MARKETS — every issuer, fund, filing, price, and event in these scenarios is fictional (macro scenarios are hypothetical). This is a workflow demonstration, not research and not investment advice. No recommendation is made with respect to any security.
● Live Custom Investigation — Enter ticker or company name:
With Claude API key: live research · Without: static demo
7-Agent Research Pipeline · Parallel Execution
Select a scenario and run the pipeline
IDLE
R
Investment Research Intelligence
SELECT A RESEARCH SCENARIO AND CLICK RUN AGENT PIPELINE TO BEGIN AGENTIC ANALYSIS
Expand a scenario on the left, then click Run Agent Pipeline
PM Research Workbench
R
Run a research scenario to populate the workbench
📄
Research output will appear here after pipeline completes
⚠
HITL authorization gates appear here after pipeline runs
📋
Audit trail will appear here as pipeline runs
🔬
Decision explanation appears after pipeline completes
⛊ BINDER VERIFIERrecomputes every SHA-256 + reproduces every valuation · runs entirely in this browser · no server involved
Drop a research-binder JSON here — or click to choose the file you exported
⚖ POLICY SANDBOXNON-PERSISTENT what-if · the active pack is never modified
ACTIVE —
SANDBOX —
Route for last-run scenario:
One rule moves → the fingerprint moves. Production discipline: policy changes ship as versioned packs; the new fingerprint stamps every subsequent ledger row. Close this panel — nothing persisted.
Synthetic research data by design (licensed market & data feeds activate in your environment during pilot) · Animation demo-compressed — see per-scenario projected production runtime (two-clock model) · Live-AI mode calls Anthropic directly from your browser; your key never reaches CAIBots.