Google Build with AI Series (Dallas) - Production-Ready AI and Multimodal Agents


Feb 12, 10:00 AM CST. Add to calendar: Google | Outlook
  • Dallas, TX (RSVP for full address) Google
  • 1 RSVPs
Description
Speaker

Join Google Cloud for the 2-days agentic AI bootcamp to learn how to build production-ready AI and multimodal agents.

Day 1: The ‘Now’ – Solving for Scale
Start building tools. Day 1 is dedicated to bridging the "Proof-of-Concept Gap"—the difficult space between a successful demo and a live, enterprise-grade application. We are moving beyond the novelty of "cool prompts" to focus on the engineering rigor required for the real world.

Day 1: What to Expect:
- Architecture for Scale: Learn how to design robust AI systems that don't just work once, but work every time.
- Data-Driven Evaluation: Replace subjective "vibe checks" with rigorous, programmatic assessment. Implement adaptive rubrics and tool-use metrics to ensure reliability.
- Security: Deep dive into enterprise-grade protection for AI workloads.

Day 2: The ‘Next’ – Capturing Innovation
The era of the text-only chatbot is evolving. Focus on the bleeding edge of AI: Multimodality. We’ll explore how to build intelligent agents that can see, hear, and respond to the world in real-time, creating immersive experiences that feel more human than ever before.

Day 2: What to Expect:
- Multimodal Gemini Agents: Coordinate agents to analyze video and audio while maintaining character consistency across multi-turn image generation.
- Intelligence Beyond RAG: Move past simple retrieval with hybrid search, context engineering, and multi-agent pipelines.
- Real-Time Live Interaction: Build low-latency, interruptible agents that "see" and "hear" using the Gemini Live API and bidirectional streaming.

Who Should Attend?
This hands-on workshop is designed for software developers, data scientists, and AI practitioners who have some experience building applications or working with models, and are looking to productionize them. To get the most out of the labs, you should have foundational knowledge of a programming language like Python and be comfortable using the command-line interface. While expertise is not required, a basic understanding of Cloud computing concepts, web APIs, and containerization technology like Docker will be highly beneficial.

To participate, you must bring your own laptop and power cable. The activities are intended for laptops and cannot be completed on a tablet or phone.

Day 1: Agenda
10:00AM - 11:00AM
Registration & Check-In
11:00AM - 11:30AM
The Production Mindset - Overview of the Google Cloud AI stack and the shift from experimentation to enterprise-grade engineering.
11:30AM - 1:00PM
Multi-Agent Orchestration - Hands-on coding session designing complex systems where multiple agents collaborate to solve tasks. Focus on state management and orchestration logic.
1:00PM - 1:45PM
Lunch Break
1:45PM - 3:15PM
Agent Evaluation - Implementing tools to measure the performance, latency, and accuracy of the agents built in Lab 1. Moving beyond "vibes" to metrics.
3:15PM - 4:45PM
Security & Guardrails
Hardening the system. Implementing prompt injection defenses, data privacy checks, and safety guardrails for agentic workflows.
4:45PM - 4:50PM
Wrap-Up & Review
Review of key learnings and Q&A with GDEs.
4:50PM - 5:50PM
Happy Hour

Day 2: Agenda
10:00AM - 11:00AM
Registration & Check-In
11:00AM - 11:30AM
The Multimodal Shift
Why text-only is the "old story." Introduction to agents that process the world like humans (audio/video/vision).
11:30AM - 1:00PM
Multimodal Orchestration
Build specialized agent crews using ADK and MCP. Master character consistency through multi-turn image generation and coordinate parallel agents to reason across text, image, and video.
1:00PM - 1:45PM
Lunch Break
1:45PM - 3:15PM
Intelligence Beyond RAG
Move past simple retrieval using RAG and context engineering for long-term personalization. Build autonomous pipelines to transform unstructured multimodal data into structured intelligence.
3:15PM - 4:45PM
Real-Time Live Interaction
Implement low-latency, interruptible agents via bidirectional streaming. Orchestrate proactive Streaming Tools and EDA for resilient, real-time control.
5:00PM - 6:00PM
Builders Demo
Rapid showcase of what was built, final resources for certification, and "Next Steps" for the community.
6:00PM - 7:00PM
Happy Hour

Venue:
Dallas, TX (RSVP for full address)

Approval Required. Your registration is subject to approval.

Contact Organizer