HUMAN EXPERIENCE DATA FOR AI
Train AI to understand what makes us human.
Real experiences. Real emotions. Real decisions.
Recall is developing consent-based datasets of how people describe real situations: what they felt, how they reacted, why they made a decision and what happened next. Structured responses, interview transcripts and original audio are planned formats, subject to explicit permissions and collection design.
Pre-launch. No datasets, samples, pricing or delivery commitments are available yet.
More than words. The context behind human decisions.
Situation
What happened, what the person expected, and the context they remember.
Reaction
What they report feeling and how they say they responded.
Decision
What they say they chose, why, and what alternatives they considered.
Outcome
What happened next, what changed, and how they reflect on it now.
Designed for teams researching human interaction.
Potential applications include conversational-agent evaluation, research on emotion-aware responses, human decision context and behavioral simulation. Suitability depends on the study design; Recall does not claim to measure internal mental states or guarantee model improvements.
Tell us what you need before we collect it.
Share the situations, languages, participant requirements, formats, volume, budget range and permitted uses you need. We'll review the brief for feasibility. Sending a brief is not a purchase agreement.
Planned output, not inventory
| Layer | Planned output |
|---|---|
| Experience structure | Context, expected outcome, feelings self-reported, action, reasoning, outcome, reflection, uncertainty |
| Source data | Original recording and/or text only where licensed |
| Processing | Reviewed transcript, collection protocol, redaction provenance |
| Rights & quality | Consent-scope evidence, documented QA, data card and limits |
We’re not selling an off-the-shelf dataset today. Submit a specification so we can evaluate interest and feasibility. A brief is not a purchase, reservation or binding quote.
Send a Dataset Brief