VAMSHI KRISHNA MACHA

I'm an AI Benchmark Engineer and AI/ML practitioner focused on agentic AI systems, LLM evaluation, and reliable production AI. I build multi-agent benchmark tasks, Python-based verifiers, Pass@K evaluation workflows, and regression-testing systems designed to measure performance and uncover failure modes. My work also spans RAG applications, MLOps/LLMOps pipelines, and computer vision for road and railway infrastructure monitoring. Recent project outcomes include improving regression detection from 61.1% to 100%, achieving mAP@50 of 0.98 in computer vision, and reducing manual QA effort by approximately 30%. I'm passionate about agentic AI, AI evaluation, trustworthy AI, open-source tooling, and building AI systems that deliver measurable results. Always interested in connecting with developers, researchers, and teams building practical AI solutions.