Inspiration
Preschool teachers see many small moments every day, but those observations are easy to lose across time and context. ChildSignal was inspired by a simple question: what if an AI agent could help teachers remember and connect those moments without labeling or diagnosing children?
What it does
ChildSignal turns a teacher's natural-language observation into structured, reviewable evidence. Capture Agent separates the activity context, invitations, observable responses, peer involvement, and follow-up questions. Teachers can edit and confirm the observation before saving it to a longitudinal timeline.
How it works
- A teacher records what happened in everyday language.
- The Strands Agents SDK powers Capture Agent to extract objective observation fields.
- The teacher reviews and confirms the result.
- Future records can be compared across time, activities, peers, and support strategies.
Why it matters
ChildSignal helps teachers notice patterns worth investigating, such as whether a familiar peer changes participation. It is designed to support professional judgment, not replace it: the system does not diagnose children or make decisions for teachers.
The demo uses synthetic data only. No real children's identifying information is used.
Built With
- amazon-web-services
- cloudflare-d1
- next.js
- strands-agents-sdk
- typescript
Log in or sign up for Devpost to join the conversation.