Cover Image for [Hands-on Workshop]From One Agent to a Swarm: Building Collaborative AI Analysts
Cover Image for [Hands-on Workshop]From One Agent to a Swarm: Building Collaborative AI Analysts
Avatar for Seattle MLOps Community

[Hands-on Workshop]From One Agent to a Swarm: Building Collaborative AI Analysts

Registration
Approval Required
Your registration is subject to host approval.
Welcome! To join the event, please register below.
About Event

What happens when AI agents stop working alone and start collaborating?


WHAT YOU'LL BUILD

Explore agentic AI through a real-world use case: multi-agent stock analysis.

Starting from a working swarm of specialized agents, you'll extend the system by adding new capabilities, introducing new agents, and shaping how they coordinate.

You'll gain hands-on experience:

  • Creating tools for agent collaboration

  • Designing handoff chains between agents

  • Orchestrating multi-agent systems

  • Building specialized agents that work together to solve complex problems

Walk away with practical experience building collaborative AI systems that go beyond single-agent limitations.


KEY TOPICS

  • Multi-agent system architecture

  • Agent specialization and collaboration patterns

  • Tool creation for agent ecosystems

  • Handoff chain design

  • Swarm orchestration strategies


WHO SHOULD ATTEND

AI/ML engineers, developers building agent systems, technical leads exploring multi-agent architectures, and builders interested in collaborative AI.


PREREQUISITES

Participants must complete setup before the workshop:

Required Software:

  • AWS account with Amazon Bedrock model access enabled (Claude Opus 4.6 via cross-region inference)

  • AWS CLI v2 installed and configured (aws configure)

  • Python 3.12+ and uv package manager installed

  • Claude Code CLI installed

Setup Steps:

  1. Add the Strands Agents MCP server to Claude Code: claude mcp add strands uvx strands-agents-mcp-server

  2. Clone the workshop repo:

git clone --depth 1 --filter=blob:none --sparse https://github.com/strands-agents/samples.git
cd samples
git sparse-checkout set python/04-industry-use-cases/finance/finance-assistant-swarm-agent
cd python/04-industry-use-cases/finance/finance-assistant-swarm-agent
  • Install dependencies: uv sync

  • Verify setup: uv run finance_assistant_swarm.py

    • Enter a ticker (e.g., AMZN)

    • Wait for the report to complete

    • If a full analysis report appears without errors, you're all set


WORKSHOP CONDUCTORS


WORKSHOP ACCESS

This workshop is exclusively available to summit attendees. First come, first served with limited capacity.

Location
The Museum of Flight
Seattle, WA 98108, USA
Avatar for Seattle MLOps Community