Aug 13, 2026 · 12 PM ET / 9 AM PT
Running multi-agent systems in production
A live look at what it takes to run multiple agents in production together — not one demo agent, but a coordinated fleet. How delegation, shared state, scheduling, and audit hold up under real load, and the operational patterns that keep a multi-agent system reliable, governed, and recoverable inside your own perimeter on Trinity.
Hosted by Eugene Vyborov
CEO of Ability.ai · creator of Trinity
Come ready to build along — 2-minute prep (optional): skim the Trinity docs and have the GitHub repo open.
Who it’s for
Built for people who ship.
Engineers running more than one agent in production and fighting coordination, shared-state, and reliability issues.
Founders and operators who want a fleet of agents that runs without them becoming the on-call SRE for all of them.
Teams weighing a sovereign, self-hosted runtime over a SaaS they'd have to send their data to.
Your host
Who’s building.
Eugene is the founder and CEO of Ability.ai and the creator of Trinity — the open-source runtime the company runs its own agent fleet on. He builds and operates multi-agent systems in production every day, live from an empty session to a deployed, autonomous fleet.
Save your seat
Join us live.
Bring your laptop and build along. Can’t make it live? Subscribe and the recording lands in your inbox.
Get it in your inbox.
A newsletter for people running AI agents in production on Trinity: hands-on guides, workshop invites, and the releases that actually matter.