I'm a software engineer. I've spent the last 6+ years building production SaaS platforms and AI systems.
Right now I'm at Klarety AI where I built the platform from the ground up. It's an AI-powered geospatial intelligence system - users ask questions in plain English and autonomous agents analyze satellite imagery for them. It's used by Yale for research, and serves hedge funds, insurance companies, and government agencies.
Some things I've worked on:
- A tiling server handling millions of requests/min - 50% faster than Felt/Atlas, cut database load by 90% with PMTiles
- LLM agent systems with rolling context summarization and semantic filtering (60% better query correlation, fewer hallucinations)
- Geospatial pipelines across 10+ satellite data sources (Sentinel-1/2, Landsat, MODIS, NAIP, Google Earth Engine)
- SaaS optimization that improved runtime by 80% and helped land a $1M contract
Previously at Aidash (SaaS for vegetation management, LLM integration for non-GIS users) and Soroco (microservices, fault detection that cut downtime by 45%).
I work across the stack - Python, TypeScript, and Go on the backend (FastAPI, Django, Next.js), PostgreSQL/PostGIS and Redis for data, Docker and AWS/GCP for infrastructure. Deep into LLM agents lately (LangGraph, LangChain, HuggingFace).
When I'm not shipping code, I'm probably reading about distributed systems, security tooling, or satellite data.