Inspiration
The first five minutes of a legal-aid intake are often spent turning a free-form story into a reviewable case record. That is where missing dates, parties, documents, and requested outcomes can delay a human decision.
What it does
CaseBridge turns one messy intake narrative into a compact, human-readable handoff:
- issue and key signals
- extracted dates, amounts, and parties
- attention signals
- open questions before routing
- suggested operational next step
- original intake preserved beside the structured output
Why not a generic chatbot?
The output follows a fixed triage contract. It exposes missing facts instead of filling them with guesses, keeps the original narrative for comparison, and marks every result as requiring human review. The prototype runs locally in the browser and sends no intake text to an external service.
How we built it
CaseBridge is a zero-dependency HTML/CSS/JavaScript prototype served by a small Python HTTP server. The current demo uses deterministic extraction so the workflow is reliable without an API key; an AI model can be added later behind the same structured output contract.
Safety boundary
CaseBridge is an intake triage aid, not a legal advice engine. It does not determine rights, liability, eligibility, or legal strategy. A qualified human must review the original material before any action.
Links
Built With
- artificial-intelligence
- css
- html
- javascript
- legal
- natural-language-processing
- python
Log in or sign up for Devpost to join the conversation.