Full Stack + AI Engineer • 7+ years crafting scalable ML/LLM systems

Senior Full Stack AI Engineer with 7+ years of experience building scalable ML and deep learning systems.

Specialized in LLMs, NLP, AI agents, and automation. I design and ship production-ready end-to-end applications (React/Next.js + FastAPI/Python + Node.js) that combine robust full-stack architecture with cutting-edge AI intelligence.

• 🔭 Currently focused on agentic AI systems (LangGraph + CrewAI + RAG) that deliver measurable business impact
• 🌱 Deep expertise turning complex AI capabilities into clean, maintainable, and scalable products
• 💼 Actively open to full-time Senior/Staff Full Stack AI Engineer roles in San Francisco or remote
• 📍 Based in San Francisco • Fluent in English & Japanese

RAG • Prompt Engineering • LLM Integration • AI Agents • CrewAI

Vercel • Databricks • EC2 • S3 • Lambda • MQTT • CI/CD

• Architected production full-stack platform using Next.js 15 + TypeScript + shadcn, FastAPI/Python, Docker + Vercel CI/CD
• Built core agentic AI layer with LangGraph + CrewAI + Pinecone RAG → delivered live importance scoring, contextual explanations, weekly digests, and real examples
• Reduced alert fatigue by 78% while shipping sub-5s alerts, visual side-by-side diffs, and GDPR-compliant Slack/Teams integration
• Helped position Parano.ai as a leader in the $1B+ agentic CI market

• Delivered production-grade UIs with React + TypeScript for real-time geospatial data visualization
• Designed scalable backend services (FastAPI + Node.js) for data ingestion and high-throughput ML model serving
• Productionized GenAI/ML pipelines on Databricks + AWS (EC2, S3, Lambda) handling large-scale climate datasets

• Built full-stack solutions with Python backend + JavaScript frontend
• Implemented high-performance MQTT pub/sub layers (Node.js + Python)
• Delivered GitHub contributions including Shell/Python utilities and CI pipelines

My best production projects are under NDA (IP-protected), but you can see my public contributions and open-source experiments in the pinned repositories below.

I actively pin repos that showcase:

• Full end-to-end LLM agent systems
• RAG pipelines with Pinecone + pgvector
• Next.js 15 + FastAPI production architectures
• AI-powered dashboards and automation tools