Infinite-scale distributed video transcoding pipeline with AWS-based queuing, multi-resolution output, and secure streaming architecture.

Cross-platform SOS platform with real-time location tracking, ML-based threat analysis, and emergency response.

Enterprise-grade hybrid RAG platform with multi-agent orchestration and tripartite storage, powered by Groq + LLaMA.

AI-powered B2B SaaS automating invoice-to-receipt matching with 98%+ accuracy via OCR and fuzzy-matching.

Splits an opaque packet stream into its components — flows, applications, and verdicts — the way a glass prism splits light. A multithreaded C++ engine with TLS fingerprinting and Prometheus/Grafana observability.

A fully working vector database built from scratch in C++, comparing HNSW, KD-Tree, and brute-force search side-by-side, with a RAG pipeline powered by a local LLM via Ollama.