February 24-26, 2027
Montreal, Canada

Agentic Engineering: Managing Agent Teams with ADK & MCP

This comprehensive workshop equips you with the skills to harness Antigravity IDE, the Agent Development Kit (ADK) and the latest Gemini models to develop autonomous voice-first agent teams using Angular.

Starting with managing agent teams and the core Agentic Loop (Plan - Execute - Verify), you will progress to building autonomous agents with goal-exit conditions, integrating MCP, A2A and Managed Antigravity Agents, culminating in a real-time Personal AI with heartbeat loops where voice intents spawn long-horizon tasks such as tickets, research and code.

Materials will be updated to the latest Gemini models.

Intro - The 2027 Shift

  • From IDE to Mission Control
  • Antigravity Platform and Gemini
  • Using Antigravity IDE as Mission Control
  • Agent Manager and Teamwork
  • AGENTS.md and SKILL.md
  • Managed Antigravity Agents

Part 1 - Managing Agent Teams

  • Managing Agent Teams with ADK
  • Code-First Development in TypeScript
  • Sequential, Parallel and Loop Orchestration
  • Lab: Boot Antigravity and Create Your First Agent Team with ADK-JS and TypeScript

Part 2 - The Agentic Loop

  • Plan - Execute - Verify
  • Working with Artifacts
  • Async Execution
  • Verifiable Workflows

Part 3 - MCP and Agent-to-Agent Communication

  • Building Tool Interoperability with MCP SDK
  • Search and Maps Grounding
  • Custom Tools and APIs
  • Agent-to-Agent (A2A) Protocol
  • Agent Cards and Discovery
  • Streaming Events and Remote Delegation
  • Lab: Constructing an Agent Team with Tool Access via MCP and A2A

Part 4 - Gemini Live and Angular

  • Developing a Gemini Live Native Audio Client with Angular
  • Angular Signals and Resource API
  • WebRTC
  • Streaming Voice and Video via WebSockets
  • Native Audio Architecture
  • Interruption Handling
  • Real-Time Transcription

Part 5 - Building a Personal AI

  • Real-Time Agent Teams
  • Heartbeat Loops
  • Persistent Memory
  • Long-Horizon Tasks
  • Lab: Building a Personal AI where Voice Intents Spawn Long-Horizon Tasks
  • Working with Tickets, Research and Code

Part 6 - Autonomous Agents

  • Autonomous Agents with Goal and Exit Conditions
  • Evals
  • Budget Guards
  • Self-Correction
  • Lab: Producing an Autonomous Agent Team that Plans, Builds and Verifies Until Done
  • Responsible AI for Autonomous Systems

Participant should have:

  • Intermediate skills
  • Basic knowledge of Node and TypeScript

Duration:

  • 2 days
  • 9:00 am to 5:00 pm
  • 1 hour lunch break included at the hotel's restaurant
  • 15 min coffee break every morning and afternoon

```

Gerard Sans

Google Developer Expert; former AWS Amplify team

Gerard Sans is a Computer Science Engineer and Google Developer Expert in AI, Cloud, and Angular. Formerly an AWS Amplify Developer Advocate, he now leads Axiom Professional Trainings and organises the NextAI London community. Passionate about the future of AI, Gerard enjoys sharing his knowledge through talks, workshops, and mentoring.

Read More

Montreal 2027 sponsored by

Become a sponsor