About Me
Cornell CS. SWE @ Google. Building AI agents, reinforcement learning systems, and applied ML infrastructure that ship. I like the layer where papers, prototypes, and product constraints collide: agent loops, evals, retrieval, and the backend glue that makes ML usable.
I'm Raj Kashikar - Cornell CS, SWE @ Google, and an AI/ML builder focused on agents, reinforcement learning, and production ML infrastructure. I've worked across Google, Cisco, AWS, Milestone, and SCMWorx; published machine learning research across IEEE, ACM, and Springer; and held research fellowships through NSF, DoD VICEROY, and AWS/Hume.
I care about the stretch between demo and deployment: agents you can trust, RL that stays stable under pressure, RAG that answers real users, and ML infrastructure that survives messy signals, latency budgets, and production constraints. The receipts are below, but the short version is 5x SWE/AI work, 3x fellowships, 8x published papers, 11x talks/posters, named grants with NSF and the U.S. Army Research Office, 5x certs across AWS, Microsoft, Intel, and Postman, and public builds on GitHub. The throughline has been cold emails, shipped prototypes, and turning small openings into internships, research, and products.
Core Work
-
Research Fellowships
3x research fellowships across NSF REU, DoD VICEROY, and AWS/Hume, with work through the National Security Institute and Hume Center on trustworthy AI, reinforcement learning, autonomous systems, RF machine learning, and national-security ML.
-
Research Talks + Posters
11x talks/posters across IEEE, ACM, Springer, AWS, Hume Center, and Commonwealth Cyber Initiative - turning RL, RFML, trustworthy ML, and AI systems work into research that can stand up in public.
-
Grants + Credentials
2x named grant researcher on NSF and U.S. Army Research Office grants, plus 5x AI/cloud/backend certifications across AWS Certified Machine Learning - Specialty, AWS Certified Data Engineer - Associate, Azure AI Engineer Associate, Intel oneAPI, and Postman Student Expert.
-
GitHub Proof of Work
Public build history across 86 repos: AI agents and LLM tooling, cloud security/MLOps, GPU/HPC plus edge inference, and recommender/RL research systems.
86 live public repos30 language footprintRepo signal clusters around agent tracing and evaluation, RAG and recommender stacks, software supply-chain security, CUDA/HPC experimentation, RF/IoT ML, quantum communication, and production-facing full-stack systems. Pinned signal includes GPU/HPC, Google Cloud supply-chain security, multimodal recommendation, and agent tracing.
