Software Engineer at Microsoft, working on Production AI, AI agents, and distributed systems. I build the systems that make AI reliable and fast for the people who depend on them.
Currently:
What I Believe
Everything I build serves someone
Infrastructure doesn't exist in a vacuum. Data quality frameworks matter because real users get wrong answers when signals are bad. Pipeline reliability matters because teams downstream are blocked when systems break. I start with who's affected and work backwards to the engineering.
Fundamentals over frameworks
System design, data modeling, algorithmic thinking, and engineering rigor transfer across any stack. I'd rather work with someone who deeply understands distributed consensus than someone who memorized an API.
The model is the easy part
Frontier models are a commodity you can swap. Production AI is won in everything around them: retrieval and grounding, evaluation, failure handling, and observability. The best AI systems are model-agnostic and reliable by design, not tied to a single provider or a lucky prompt.
Boring infrastructure wins
The best systems are invisible to the people who depend on them. They serve millions of requests, degrade gracefully, never lose data, and nobody has to think about them. That invisibility is the product — for data platforms and AI systems alike.
Get in Touch
Open to conversations about engineering roles, data infrastructure, and AI systems at scale.