Hands-on AI workshop: Multimodal agents
summary
90 minutes of hands‑on steps show how to build and deploy multimodal agents that see, hear, and reason: code clones and Colab link, Vertex AI models (Gemini / Nano Banana 2 for images, Veo 3.1 for short video), BigQuery star catalog, FastMCP + Google‑managed MCP tools, ADK agents and Web UI, and Cloud Run deployment via Cloud Build. The workshop generates a portrait + circular icon, runs a multi‑turn image generation with Gemini, runs three agents in parallel and locks a location with parallel orchestrator with two‑of‑three consensus, and demos Cloud Run deployment via Cloud Build — everything runs end‑to‑end.
AI-written summary · about
chapters (14)
- 0:00Introduction
- 3:42Environment setup
- 9:37Clone repository and install
- 16:13Customize avatar
- 21:12Build avatar generator
- 33:39Generate identity
- 36:45Level 1 setup
- 39:22Evidence generation
- 45:24Star catalog setup
- 46:59Build MCP tools
- 50:26Deploy MCP server
- 53:29Build agents
- 58:07Finalize workflow
- 1:00:17Test and conclude
related talks
01h 01mBuilding Multimodal Apps with Gemini 2.0 and Gemini SDK
37mBuilding Multimodal AI Agents From Scratch
44mHands-on-lab: It's all about going hands-on with Generative AI!
25mBuilding AI Agents
46mMultimodel Multimodal and Multiagent innovation with Azure AI
39mBuild the new generation of AI Agents
30mLook ma, no hands! Multimodal AI-Agents in the browser
16mCreate agent skills for on-device generative AI