01
Meera
BuiltLocal AI operating layer that plans, executes tools, and verifies outcomes instead of trusting the model’s word.
Local AI · Agents · LLM/Node.js · Ollama · Qwen 2.5 · SearXNG · Playwright · SQLite
Original work
02 / Projects
A collection of AI systems, experiments, and engineering work. Status and ownership are explicit. Nothing here is padded with metrics that were not measured.
5 projects in All
01
Local AI operating layer that plans, executes tools, and verifies outcomes instead of trusting the model’s word.
Local AI · Agents · LLM/Node.js · Ollama · Qwen 2.5 · SearXNG · Playwright · SQLite
Original work
02
Also KRAVEN
Dependency security analysis that turns OSV findings into a brief, then patches only after explicit approval.
Security · Automation · LLM/Python · LangGraph · Gemini 2.5 Flash · OSV · osv-scanner · Rich
Original work
03
Hybrid RAG for support questions: vector search plus BM25, fused, then answered by a local LLM with citations.
RAG · LLM · Local AI/Python · LangChain · FAISS · BM25 · sentence-transformers · Ollama
Original work
04
Serverless Discord study companion: slash-command in, Groq response out, no long-running gateway process.
LLM/Python · Flask · Discord Interactions · Groq · PyNaCl · Vercel
Original work
05
Local-first personal AI with memory and agent orchestration — studied from existing open-source work, not claimed as original.
Local AI · Agents/Rust · TypeScript · SQLite · Ollama · Tauri
Based on existing open-source work