-
LLM Function Calling Explained: How to Connect AI to Your Real APIs
I had a chatbot that could answer questions about our product catalog. It was helpful — until someone asked “what’s the current stock for yoghurt in store 1234?” The model confidently hallucinated a number. It had no way to check. It just guessed. The fix wasn’t a better prompt. It was giving the model the ability to call my…
-
Graph-as-Semantic-Layer: Why Your API Architecture Is Your AI Strategy
In my previous post on AI agents, I ended with a question: what if your entire API surface was already typed, self-describing, and machine-readable? What if the agent didn’t need hand-written tool descriptions at all? I’ve been thinking about this ever since I spent weeks deep-diving into GraphQL Federation for a large retail application. The architecture I…
-
What Is an AI Agent? How LLMs Go From Chatbots to Taking Actions
You’ve built a chatbot. It answers questions. It summarizes documents. It’s useful — but it’s passive. It waits for input, generates text, and stops. It doesn’t do anything. Now imagine something different: you tell an AI “find all products expiring within 48 hours, check their stock levels, and create markdown recommendations for anything with more than 10 units remaining.” And it does it. Not by generating a…