Features
Built for every research use case
From quick concept tests to multi-party simulations and reusable studies, person.run helps teams run research workflows with transcript-level visibility.
Build personas from demographic seeds
Define age, income, location, education, values, media diet, and life stage. Person.run generates a fully realized individual with coherent opinions, speech patterns, and decision-making logic.
- Natural-language persona description editor
- Demographic seed inputs (age, income, location, psychographics)
- Timeline memory model for consistent life history
- Consistency checking to catch contradictions
- Preview mode with live sample responses
Reusable studies with run history
Save one population + one skill + one context as a reusable study, then rerun on demand with full run tracking and export-ready outputs.
- Reusable study definitions with configurable context
- Run history with queued, running, completed, and failed states
- Optional participant subset overrides per run
- Collated exports in JSON, JSONL, and CSV
- Per-run context and web-access snapshots for reproducibility
Turn-based debate, negotiation, and collaboration
Configure optional multi-party skill behavior for rule-driven sessions where personas take turns, react to prior turns, and produce structured outputs.
- Turn policies with deterministic participant ordering
- Role-aware prompts and shared/per-participant output modes
- Turn-by-turn transcript capture across rounds
- Pause, resume, and cancel controls during execution
- Automatic fallback to single-party runs when multi-party is not enabled
Turn transcripts, traces, and scores
Inspect every turn, monitor run-level traces, and evaluate quality outcomes with exportable results for downstream analysis.
- Turn transcript viewer for study and population runs
- Sentiment and theme analysis across responses
- Demographic breakdowns and cross-tabulations
- Per-run traces and scoring diagnostics
- Per-persona response history and consistency scores
- JSON, JSONL, and CSV exports for downstream pipelines
Fits your existing workflow
Person.run is API-first with an OpenAPI spec, typed SDKs, managed automation workflows, and webhook support. Pipe persona responses into your survey tools, CRMs, or analytics platforms.
- Full REST API with OpenAPI 3.1 specification
- TypeScript, Python, and R SDKs
- Endpoints for studies, run turns, and exports
- Managed automation templates for Zapier, Make, and n8n
- Webhook callbacks for async API job results
- External executor callbacks for provider orchestration flows
- Managed destination test sends and replayable delivery logs
- Document upload for context-grounded responses
- Batch APIs for high-volume panel runs
Collaborate across your research team
Clerk organizations, role-based access, and workspace-scoped controls let teams collaborate safely without losing tenant isolation.
- Organization switcher for multi-workspace teams
- Tenant-scoped workspaces with data isolation
- Enterprise SAML SSO for organization members
- API key creation with scope, role, and expiration controls
- Per-tenant usage limits and quota enforcement
- Role-based access control (Owner, Admin, Editor, Viewer)
- Append-only audit log with dashboard viewer
And more
Everything else you expect
Turn transcripts
Round-by-round transcript retrieval for population and study runs.
50+ languages
Run studies in any language with personality-preserving translation.
Model-aware controls
Per-request model selection with graceful fallback warnings and audit metadata.
Transparent billing metadata
Prompt settlement includes flat per-tool invocation billing with modifier attribution for awareness and mood.
Encryption
All data encrypted at rest (AES-256) and in transit (TLS 1.3).
Timeline revisions
Revise and supersede timeline memories with full revision history.
Async study pipeline
Queue large panel runs with status polling, plus callbacks for async API jobs.
Ready to put these features to work?