11 stories
Most explanations of agentic AI systems stay at a high conceptual level. In this video, I trace a single user request step-by-step through a production-level agentic AI state machine from start to finish. I break down the three core components that make AI agents deterministic and reliable: 1.
Before you write a single line of LangGraph, CrewAI, or Python code, there is one control flow graph you must draw as part of your system design. In this video, I share a 5-question framework to help you architect reliable agents, protect irreversible actions, and decide whether you need an…
Your company just paid out a $200 refund on a $100 order. You look at the codebase and nothing is broken, nothing failed, the model didn't hallucinate. This is a classic orchestration failure.
Why do most multi-agent AI systems fail in production? Learn how the A2A (Agent-to-Agent) Protocol works with the help of a complete architecture walkthrough.
Model Context Protocol (MCP) promises to simplify how AI Agents interact with tools. But what actually changes when you use it in a real system? In this video, I take an existing multi-agent security application built with Red, Blue, and Judge agents, and refactor it using FastMCP in Python.
The industry is currently obsessed with multi-agent architecture, but scaling your agent count drastically increases your coordination tax, and the failure rate compounds.
Everyone wants to build AI agents - but underneath every agent is one raw API call most people never look at. A short, no-frameworks look at what actually happens when you call an LLM API - the raw request, the raw response - that's it.
Most people think LLMs are just for chatting, but tool calling turns them into agents that can actually do work. Function calling is the "bridge" that allows AI agents in Generative AI to move beyond simple chat and actually perform actions in the real world.
Your AI agent works perfectly. Until it refunds $20,000 instead of $2,000 at 3am, with no human watching. Or a scammer crafts a message that tricks it into sending money to a fraudster. No amount of self-critique catches this.
When does the AI Router pattern fail? Today’s system design breakdown covers the Pipeline and Parallel scaling patterns for Multi-Agent Systems. We map out how to chain dependent agents, how to run independent agents simultaneously, and the exact architectural limit where both of these hardcoded…
Mapping out the foundational architecture for modern Agentic AI systems. Today's System Design breakdown: The structural difference between a standalone LLM and an Agentic Control Loop.
Showing 11 of 11 stories
You're all caught up.
Get the top AI news and tools straight to your inbox every week.
Weekly · AI news from The One AI Architect
Join 15,000+ AI enthusiasts
