I build software systems that solve messy, real-world problems and hold up in production.
Most of my work has been around production AI systems, data pipelines, and automation. I currently work on conversational AI agents for auto dealerships, where the focus is not demos or prototypes, but evaluation, response quality, cost, and latency in real operating environments. Shipping is the easy part. Making systems reliable, debuggable, and economical at scale is the real work.
Before and alongside this, I have built end-to-end data ingestion pipelines, including browser automation for legacy systems that do not expose reliable APIs, large-scale scraping, and inventory normalization for downstream use. I have also worked on core platform components such as CRM systems, workflow orchestration, and internal tooling that supports these products.
I enjoy working at the intersection of ambiguity and constraints. That often means dealing with bad data, partial integrations, unclear workflows, and systems that were never designed to scale. My approach is to simplify aggressively, design for failure, and optimize for long-term maintainability rather than quick wins.
I am particularly interested in how AI can be applied pragmatically to existing workflows. Not as a layer of hype, but as infrastructure that improves reliability, speed, and decision-making. I also care about how teams build, and have spent time adopting and integrating modern AI-assisted development tools to improve developer velocity without sacrificing code quality.
At a fundamental level, I like taking complex systems, understanding how they actually behave in the real world, and turning them into software that people trust and use every day.