Raghunandan Madishetty

I've contributed to OpenSRE-Agents, implementing core CLI subprocess management and autonomous agent execution features that improved execution workflows and system reliability.

I also collaborated on an Agent Monitoring module that improved observability and reduced manual monitoring effort by 33%, while contributing GitHub pull requests, code reviews, fixes, and improvements using Python and Azure.

I've built full-stack AI products including ChessAI, where I created a two-layer chess move validation system and integrated fine-tuned SLM move suggestions with server-side validation. For LearnlyGPT, I built document-processing APIs and RAG pipelines that processed 140+ documents and reduced average query latency by 64%.

My work also includes an emotion-recognition pipeline trained on RAVDESS audio data, where feature engineering and augmentation raised classification accuracy from 63.0% to 82.27%. I enjoy applying Python, React, FastAPI, PyTorch, and modern AI tooling to practical systems.