ConvoPilot
AI-powered conversational agent built for a hackathon — awarded 2nd runner-up. Uses local LLMs via Ollama and a RAG pipeline over Qdrant for grounded, context-aware responses.
Senior Software Engineer
6+ years shipping Python backends, RAG pipelines, and IoT systems — from GSoC at Apache to leading AI product development at a Tokyo-based startup. Currently exploring the overlap of Gen-AI and reliable distributed systems.
AI-powered conversational agent built for a hackathon — awarded 2nd runner-up. Uses local LLMs via Ollama and a RAG pipeline over Qdrant for grounded, context-aware responses.
Multi-tenant retrieval-augmented generation platform serving a Japanese AI product startup. Handles document ingestion, semantic search, and LLM-orchestrated Q&A at production scale.