At Samsung R&D Institute, I contribute to the Tizen OS application framework powering Amazon Prime Video and Netflix. I also resolved production-critical OTT crashes and designed an MQTT-based Device Automation Bus that reduced manual testing effort by approximately 40%.
I implemented semantic caching in the Rust-based TizenClaw AI agent using ONNX Runtime, reducing retrieval latency by up to 50% for repeated and similar queries. I also built a proof of concept integrating Swiggy MCP servers with TizenClaw for natural-language food discovery and ordering.
Earlier, as a Software Research and Developer Intern at Samsung R&D Institute, I built a LangChain-based agent to automate GTest generation for C++ projects, reducing manual effort by 60%. Outside work, I developed a full-stack vintage fashion marketplace using Django and PostgreSQL.
