Gemini Tales: From ADK Tutorials to a 5-Agent Storyteller

Hi everyone! Just submitted Gemini Tales — an AI storyteller that doesn't move until the child does.

This was a brutal but rewarding journey. Even with my background, scaling from initial ADK concepts to a full multi-agent system with A2A protocol on Cloud Run was a massive deep dive.

Thanks to the creators of the "Way Back Home" course in the MultiModality Agents resources — it completely changed my architectural approach 4 days before the deadline!

I also want to thank the organizers for the incredible opportunity to be among the first to dive into this new era of See, Hear, Speak agents. Building with Gemini’s native multimodal capabilities was a game-changer.

I'd love to hear your thoughts on fighting sedentary behavior with AI.
Check it out here: https://devpost.com/software/gemini-tales

Good luck to everyone!