

Build Research Agents That Don’t Break: LangGraph + Flyte
Build a research agent that actually works in production. In this hands-on workshop, we'll wire up a ReAct agent pattern using Tavily for web search, LangGraph for reasoning and tool-use, and Flyte 2 for durable orchestration that scales. By the end, you'll have a working agent with a Gradio UI that runs on your laptop or scales across a production Flyte cluster.
What we'll cover
A practical intro to AI agents and the ReAct pattern
Tavily: web search built for LLMs, not humans
LangGraph: composing the agent's reasoning and tool loop
Flyte 2: orchestration, durability, and scale-out across workers
A Gradio UI for interacting with your agent
Patterns for extending the agent to your own domain
What you'll leave with
An end-to-end research agent you can run locally today
A clear path to production if you have Flyte cluster access
A portfolio-ready project and a certificate of participation
Who it's for
Engineers and AI practitioners who want to move past toy agent demos. Whether you're building at work, learning something new, or just want to ship something real, you'll leave with code you can keep extending.
Hosted by Sage Elliott, AI Engineer at Union.ai