> I architect and build scalable backend systems that bridge the gap between robust enterprise applications and cutting-edge AI.

> Currently at Therap (BD) Ltd., I focus on AI-native architecture, system optimization, and production-grade reliability. I specialize in designing context-aware AI workflows, LLM-powered data assistants, and high-speed data pipelines using Java, Spring Boot, ClickHouse, and LangGraph. My recent work includes implementing custom context-fetching protocols via MCP to accelerate developer productivity, and designing zero-leakage SQL security layers for multi-tenant architectures.

> Beyond my day-to-day enterprise work, I am deeply invested in context engineering and agentic workflows. I actively build tooling that centralizes instructions and contextual memory for AI assistants into project-specific directories, and I enjoy exploring multi-agent system designs.

> Whether it is modernizing a legacy monolith or building an intelligent, multi-agent platform from the ground up, I love solving hard engineering problems. If you are building systems that need to scale intelligently, let's connect.