← back
Agents in Production: How OpenGov Built and Scaled OG Assist - Gabe De Mesa, OpenGov
Takeaway
A shared agent platform scales across products when domain tools, runtime control, contracts, and feedback are designed together.
Summary
- OG Assist embeds a shared assistant across OpenGov ERP products, with product teams contributing tools and skills for their domains.
- The assistant can query suite data and inspect or highlight on-screen elements to guide users through workflows.
- OpenGov replaced LangGraph with an Effect-native TypeScript loop for tighter control over tracing, concurrency, error handling, and model injection.
- A2A provides a shared frontend/backend contract, while user feedback and CI evaluations check responses and tool use.
opengoveffectproduction-agents
Original description
Come and learn about building AI Agents in production. Learn hands-on directly with the AI Agents team from OpenGov which powers AI workflows across thousands of state and local governments. This session will cover: The core agent loop/harness A2A protocol Building with Effect-TS and Typescript Feedback and evals Long context handling Monitoring and observability Building out tools and skills Enterprise contribution model Accelerating workflows with Claude and Cursor Speakers: Gabe De Mesa (OpenGov): Gabe works on the flagship AI Agent product offering at OpenGov which serves thousands of state and local governments across the country. Gabe was one of the first engineers to join the newly formed AI Agents team. He wanted to share the features he's built, experience he's gained building and running agents in production, battle scars gained from running with real production workloads and customers, architecture, and some best practices to help other engineers run agentic workloads at their companies. X/Twitter: https://x.com/jamesjellow LinkedIn: / gabedemesa GitHub: https://github.com/gabedemesa