Vijaydeep·Aug 15Loop Engineering: Building Reliable AI AgentsFrom ReAct to purposeful loops: how feedback and evaluation make agents more reliable
Vijaydeep·Aug 2Why ReAct Alone Isn’t Enough: The Next Evolution of AI AgentsReAct taught AI agents to reason and act. The next challenge is ensuring those actions actually lead toward the right goal.
Vijaydeep·Jul 2Why ReAct Changed Everything: From One-Shot Reasoning to Intelligent ExecutionUnderstanding the architectural shift that transformed language models into AI agents.
Vijaydeep·Jun 29Why AI Agents Still Fail: The Missing Feedback LoopReasoning and tool calling made AI agents more capable, but not necessarily more reliable. In this article, we explore why one-shot…
Vijaydeep·May 7Skills over MCP — Shipping Business Knowledge to AI at RuntimeEarlier we covered how a single gateway can onboard any number of backends with zero integration code. If you haven’t read Part 1, it’s a…
Vijaydeep·Apr 26One MCP Gateway, N Services, Zero Integration CodeHow to scale an MCP gateway with your API estate — not your engineering team.
