The Human in the Loop

Someone has to keep the agents in line

✨ README

I'm Anup, an AI Engineer and Software Architect. I design and deploy Generative
AI and agentic AI platforms, with a focus on production-grade architectures
for enterprise and defence contexts.
My work spans multi-agent orchestration, durable workflows, RAG, and LLM
application engineering. I use tools like LangGraph, Temporal, and LlamaIndex,
and emerging protocols like MCP and A2A. I share practical lessons from the
field through The AI Engineering Brief.

🔭 What I'm working on now (2026)

• Designing agent orchestration patterns with LangGraph and Temporal for defence, finance, and use cases with regulated workflows
• Exploring MCP and A2A as a foundation for cross-organisation agent communication. Convinced this is where the next layer of the AI stack lives.
• Writing a series in *The AI Engineering Brief* on running multi-agent systems in production: failure modes, observability, retries, and the unglamorous middleware nobody demos.

If you're building in this space, I'd like to compare notes. Feel free to reach out.

But it’s precisely by knowing that our knowledge is limited that we are able to learn. Between absolute certainty and ignorance there’s all this interesting space in which we live. - Carlo Rovelli