Building AI agents end-to-end: A hands-on workshop

This workshop provides a structured, hands-on journey through the end-to-end process of building AI agents using Strands Agents, an open source agent harness SDK.

Participants begin by constructing a foundational agent and progressively learn about various advanced capabilities like configuring model providers like Ollama, integrating various services, adding MCP tools, and streaming response handling. The workshop also addresses critical production considerations such as implementing safety guardrails, enabling persistent memory for contextual conversations, and adding observability and evaluation frameworks. By the end of the session, participants will have built a fully functional, production-ready AI agent and gained practical experience with the core patterns and best practices for agent development.

Workshop materials - GitHub Link

Workshop Infrastructure

Attendees will be provided with the cloud infrastructure and access to cloud models to run the workshop.

Duration

2-3 hours

Attendees will learn:

1. How to configure AI agents and swap between cloud or local foundation models.
2. How to write custom Python tools enabling agents to autonomously execute actions and API calls.
3. How to add external knowledge and securely integrate live data using the Model Context Protocol (MCP).
4. How to implement strict guardrails to prevent hallucinations, secure data, and enforce structured outputs.
5. How to persist memory and transition from stateless scripts to stateful agents with long-term databases.
6. How to add observability, trace agent reasoning, monitor token costs, and evaluate performance.
7. How to perform advanced orchestration and connect specialized agents into complex multi-agent workflows.