Understanding AI Agents in Camunda

What to expect

Go beyond the BPMN demo to dive "under the hood" and into the technical details of how AI agents work inside Camunda.

Watch as Developer Advocate Nathan Loding walks through deploying our AI Email Support Agent available in Camunda Marketplace and goes into the technical details of each step. What you'll learn:

• Prompt analysis: how the agent interprets and acts on incoming messages.
• Data storage and retrieval: where variables and memory chunks are stored, and how they're used during orchestration.
• Vector database integration: how Camunda Embeddings Connector enables long-term memory and Retrieval-Augmented Generation (RAG).
• Guardrails and escalation paths: how BPMN handles things like sentiment thresholds or token limits to escalate a human reviewer.
• Ad-hoc sub-processes: how incorporating AI agents into these can support dynamic, scalable automation.

If you're building with agentic AI patterns, watch the recording of this live session for a concrete example that will give you the understanding you need to get started with real components, connectors, and code.