Agent · Synthetic populations
SurveyLM
Simulate decisions and behaviour with synthetic agents calibrated on your data.
SurveyLM lets you pre-test policies, communications, and trade-offs before human pilots — or run one final pre-rollout check. Agents are calibrated on the data you already have and probed with the same instruments you use in field work.
Key features
- Dynamic scenario sweeps across populations and contexts.
- Contextualised Q&A with calibrated agents.
- ALM feedback integration to refine framings.
- Pilot-style runs without the cost of recruiting humans.
Status: Live demo · seeking trial partners