Durable, composable and scalable agentic engineering with spring-ai-agent

Event details

Date

Thursday, 3 December - 19:00

Place

Sonar

Genre

Conferences

Organiser

GenevaJUG

Description

Durable, composable and scalable agentic engineering with spring-ai-agent With Sébastien Deleuze Getting an LLM to call a few tools is the easy part. Running agents in production is harder: a multi-step agent may crash partway through, pause for human input, hand work to other agents, and stream tokens to thousands of users at once. spring-ai-agent is a new Spring AI module for these challenges, built with the same type safety, extensibility, and operational focus as the rest of the Spring portfolio. Composable: an agent is a typed Agent transformation. You can build one from an existing ChatClient or from a graph whose nodes call the model and run tools. Graph agents can orchestrate nested agents, with conversations that can be shared, forked, or isolated. Mismatched contexts are caught when the agent is built, not halfway through a run. Durable: the durability feature snapshots each completed step, so a run that crashes resumes where it stopped instead of calling the LLM again. The same mechanism supports human-in-the-loop workflows: interrupt nodes pause a run and resume it with a human’s answer, even deep inside nested agents. Observable and extensible: logging, conversation persistence and compaction, and durability are all pluggable features. Each is written once on a single set of hooks and sees every model call, tool call, and node, including those of nested agents. Scalable: spring-ai-agent provides a unified API for either a blocking call or an imperative event stream with backpressure, designed for virtual threads and structured concurrency, with no reactive programming required. You’ll leave knowing how to design agentic systems that are reliable, testable, and ready for production, and how the new module fits into the Spring AI programming model.