- Published on
A local-first RAG system that uses open-source LLMs, embeddings, and vector search to answer questions across legal documents with grounded source citations.
All Posts
All Posts
- next-js (8)
- logging (2)
- observability (2)
- node-js (2)
- tailwind (2)
- hcaptcha (2)
- resume (2)
- ai (1)
- rag (1)
- local-ai (1)
- ollama (1)
- langchain (1)
- embeddings (1)
- vector-database (1)
- vector-search (1)
- similarity-search (1)
- legal-tech (1)
- python (1)
- uv (1)
- docker (1)
- chroma (1)
- gemma (1)
- gemma-3-12b (1)
- nomic (1)
- nomic-embed-text-v1 (1)
- pino (1)
- react (1)
- typescript (1)
- animation (1)
- homepage (1)
- languages (1)
- recaptcha (1)
- security (1)
- privacy (1)
- contact (1)
- blog (1)
- website (1)
- vercel (1)
- email (1)
- resend (1)
- brevo (1)
- mdx (1)
- contentlayer (1)
- contentlayer2 (1)
- content-collections (1)
- turbopack (1)
- webpack (1)
- middleware (1)
- proxy (1)
- uuid (1)
- pdf (1)
- react-pdf (1)
- Published on
How I added structured server-side logging with Pino, request-scoped loggers, redaction, and consistent route/service logs in a Next.js app- Published on
Using the Next.js proxy middleware pattern to add per-request IDs, unique visit IDs, and a composable middleware context- Published on
A breakdown of the animated Hello World component on my homepage.- Published on