In this project, I build a Fitness Coach AI Agent end-to-end using an agentic workflow (AGNO + Gemini Flash) that can generate a personalized workout plan and a meal plan based on your body type, goals, and nutrition needs. The core idea is simple: instead of a generic “one-size-fits-all” template, the agent takes your inputs (like goal, activity level, preferences) and then reasons through what a good plan should look like—training split, progression, calories/macros, and meal structure—before it outputs a clean, actionable plan. I also connect this build to the agentic design patterns I teach in my prerequisites: how to think in terms of roles, tools, and structured outputs so the agent doesn’t hallucinate random fitness advice. You’ll see a full demo of the agent generating different plans for different individuals, plus a step-by-step path to replicate it on your machine. If you want the code, I’ve shared it on my GitHub, and if you want deeper guidance, I also link my Medium (for write-ups) and Topmate (for 1:1 calls) so you can move faster on real GenAI system design projects.