At Caterpillar, via COMTEK International, I lead engineering for Cat AI Assistant, a unified enterprise assistant connecting fleet management, knowledge bases, dealer interactions, and e-commerce workflows.
I designed and implemented its multi-agent LLM system with LangChain and LangGraph, alongside a Go and gRPC tool execution layer. I also developed backend systems for context aggregation, tool routing, and conversational state management.
I implemented RAG pipelines across enterprise knowledge systems and integrated Azure OpenAI, AWS Bedrock (Anthropic), and Gemini. I built observability and evaluation pipelines with Langfuse, including metrics for response quality, latency, and tool execution success.
Before this work, I built data pipelines and MLOps workflows at Caterpillar and developed machine learning models and analytics solutions at Caterpillar, Severstal, and Whirlpool EMEA. My background spans data science, backend systems, and production AI engineering.

