Travis User

I am a results-driven LLM engineer and AI Ops specialist with 8+ years designing, deploying, and optimizing production-grade Generative AI and LLM systems. I focus on LLMOps, prompt engineering, observability, and edge deployment.

At Dell Technologies I built prompt evaluation pipelines using PromptFlow and LangChain that reduced application development time by 85% and increased marketing conversions by 30%. I integrated human-in-the-loop feedback loops, designed a custom observability suite (Splunk + summarization models) that cut incident resolution time by 60%, and launched FastMCP-powered agents that improved sales targeting efficiency by 40%.

Earlier I deployed ML models across supply chain infrastructure, engineered predictive models that helped prevent outages and saved roughly $5M quarterly, and developed tooling that reduced data onboarding time by 90%. I have contributed to open-source projects and hold a patent for a machine-learning based database recommendation system.

I am completing an MSc in Data Science with a thesis on LLM summarization for resource-constrained devices and work regularly with LangChain, HuggingFace, vLLM, Llama.cpp, and related toolchains. I seek to build explainable, human-aligned AI pipelines that deliver measurable business impact.