Bring a laptop. Leave with a prototype. GDG Tulsa hosts an in-person Build with AI session on May 3 built as a working session, not a lecture. We open with a tight walkthrough of the Google AI tech stack — Gemini, AI Studio, Vertex AI, prompt-to-app workflows, and lightweight agent patterns — then shift the room into build mode. Attendees pick a track (chatbot, document Q&A, multimodal demo, or workflow automation), get a starter scaffold, and ship something that runs by the end of the night. Organizers float the room for live debugging, API setup, and architecture calls. We cover the practical stuff most tutorials skip: API key hygiene, rate limits, when to switch from AI Studio to Vertex, deployment to Cloud Run, and how to package a demo you can actually show off the next morning. The session closes with a round of two-minute show-and-tells from anyone who wants to demo what they built. Attendees leave with working code, a deployed endpoint, and a clearer picture of which corners of the Google AI stack fit their next project.