Kobby Owen

Built AI-driven agents to process and analyze 100K+ semiconductor data points, surfacing metrics that reduced analysis time from days to hours. Owned backend APIs, authentication, and platform services, and contributed to frontend experiences and product delivery. Improved internal engineering workflows with automation, cutting manual effort by 40% and turnaround time by 45%. Resolved critical production issues across AI infrastructure, improving reliability, security, and platform stability, and implemented observability, metrics, and monitoring for stronger operational insight.

I built scalable Python pipelines for large-scale LLM evaluation across 500K+ prompts, designing structured JSON schemas and automated frameworks for consistency, accuracy, and reliability testing. I also created high-quality training datasets through advanced data parsing and worked closely with researchers to integrate human feedback loops to improve model alignment.

Engineered scalable Python pipelines for large-scale LLM evaluation workstreams. Built structured evaluation frameworks using JSON schemas to validate model quality, consistency, and reliability. Developed benchmarking systems to measure accuracy, latency, and robustness. Created preprocessing pipelines using BeautifulSoup, regular expressions, and structured datasets.