Inspiration

Inspired by tools like Claude Code that can ingest and work with any type of data through agents, we wanted to bring that same power to AI litigation research. The DAIL database is rich but inaccessible — we set out to change that.

What it does

OpenDAIL lets you explore AI litigation cases using plain English. Ask a question, get instant charts and insights. It uses custom agents to convert natural language into SQL, execute queries, and return interactive visualizations — no technical knowledge required.

How we built it

We started by building the frontend and backend in isolation, developing each side independently before connecting them together. The backend uses FastAPI powered by Ollama LLM for natural language to SQL conversion, with SQLite storing the DAIL case data. The frontend uses React, Recharts, and Tailwind CSS to render dynamic charts and a conversational chat interface. Custom agents handle the full pipeline from query to chart.

We used AI coding tools including OpenCode and Cursor alongside writing code ourselves, and leveraged Claude, Gemini, and ChatGPT for research throughout the process.

Challenges we ran into

Creating a universal response type that worked across all query types was tough. We also had to design a reliable agent workflow so tool calling was fast and accurate, while keeping the model grounded in the database to prevent hallucinations and ensure answers were always backed by real data.

Accomplishments that we're proud of

We successfully built the agent pipeline we envisioned from day one. We're also proud of the breadth of support we added — multiple chart types, extensive filtering options, and a clean interface that makes complex litigation data genuinely approachable.

What we learned

We learned how to build systems that hold up under large amounts of unstructured data, and how to design agent workflows that turn that chaos into structured, meaningful output reliably and efficiently.

What's next for OpenDAIL

We want OpenDAIL to become the go-to visualization layer for the DAIL database. Next up: expanding the agent ecosystem with case feedback agents, stats agents, and more — making OpenDAIL a full research companion for AI litigation.

Built With

Share this project:

Updates

Submission history