AI / ML

  • RAG pipelines
  • FAISS
  • SentenceTransformers
  • llama-cpp-python
  • AutoGluon
  • Streamlit
  • GGUF quantization

Languages

  • Python
  • Rust

Tools & Other

  • Git
  • Docker
  • Google Sheets API

I build AI systems that run entirely locally — no cloud, no API keys, no data leaving your machine. My AI-Companion project is a modular local assistant with long-term semantic memory: RAG pipeline using FAISS and SentenceTransformers, running quantized GGUF models via llama-cpp-python.

I care deeply about privacy-preserving AI. The idea that every interaction needs to go through a third-party API feels wrong when modern hardware can run capable models on-device. I've spent significant time on GGUF quantization, memory-efficient retrieval, and making local inference actually practical.

I also built auto-ml, a Streamlit app wrapping AutoGluon to make ML model training painless — upload a dataset, pick a target, let it run. It's the kind of tool I wish existed when I started with ML.

currently exploring

Local LLM inference optimization, semantic memory systems, and making on-device AI genuinely useful for real workflows.

Projects

Google Cloud Badges

View All Badges →