Havrus

Havrus comes from a problem I saw up close as a lawyer and as a teacher of electronic court filing: legal work repeats a lot of what a machine does better, and the tools on the market were built by people who never filed a petition.

The product is a RAG-based AI engine for law firms: case-law analysis, identification of arguments and risks, and drafting support. Multi-tenant architecture with privacy (LGPD) by design. It's where my two backgrounds meet.

Java with Spring Boot on the backend and Angular on the frontend, the stack my project partner and I are most productive in. AI is part of the product rather than a layer on top: language models with retrieval (RAG) over the firm's own material, so answers stay anchored in the right documents.

In progress. This case study will grow with architecture decisions, results and what changed along the way as the product moves forward.