RAG · LLM applications · AI agents · retrieval systems · real-time software
I started by building production games and real-time multiplayer systems. Today I'm focused on AI applications where retrieval, orchestration, evaluation and software engineering matter as much as the model itself.
AI-first projects, followed by systems built in production.
AI-powered YouTube gap analyzer that researches content, identifies opportunities and grounds analysis in retrieved evidence.
PG management platform for owners and tenants, with an AI layer that uses tool calling to work with live structured application data.
Live multiplayer strategy game. Worked on client-server state, real-time synchronization, alliances, progression, state machines and performance.
Production real-money gaming applications across Ludo, Rummy, Snake & Ladder and slot games, with deterministic game-state and mobile performance work.
An immersive VR project from the same production environment, extending the real-time 3D side of my engineering background.
I care about the engineering around the model: retrieval quality, orchestration, APIs, caching, evaluation and reliable execution.
LLM Applications · RAG · AI Agents · LangGraph · Tool Calling · Evaluation
Vector Search · ChromaDB · BM25 · Hybrid Retrieval · Semantic Search · Reranking
Python · C++ · C# · APIs · Redis · SQL/NoSQL · System Design · DSA
Unity · Unreal Engine · Multiplayer Networking · Photon · Socket.IO · Optimization
Teach Python, DSA, ML, deep learning, AI fundamentals and software architecture. Mentor 400+ students through technical projects, debugging and implementation quality.
Owned architecture and delivery of Warrior of the Wasteland. Designed client-server models, synchronization, resource/progression systems and concurrent multiplayer sessions.
Built production mobile gaming systems in Unity/C#, including Ludo, Rummy, Snake & Ladder and slot games. Profiled and optimized C# systems for mobile hardware.