Hosted by the Agentic AI Foundation

Voice agents are moving past demos, and the production problems are showing up fast.

Users feel latency straight away. A pause feels broken. A missed interruption feels rude. A bad handoff damages trust. A polished demo can still struggle with real users, noisy environments, edge cases, and operational constraints.

Voice Agents Forum is a one-day event for the teams building and deploying voice agents in production.

It is for the engineers, founders, product leaders, researchers, and operators working through latency, turn-taking, interruptions, orchestration, evaluation, observability, compliance, reliability, and human handoff.

About the event

Voice Agents Forum focuses on practical lessons from teams deploying voice agents across customer support, healthcare, financial services, enterprise operations, and consumer applications.

The aim is to surface the details that usually get missed in public talks: what broke after launch, where architecture decisions mattered, how teams approached evaluation, and what changed once reliability, compliance, and trust became production requirements.

Expect practical sessions from people building and operating voice systems in the real world, with a focus on what worked, what failed, and what teams would do differently next time.

Hosted by the Agentic AI Foundation and brought together with the MLOps Community.

Agenda

9:00 AM - 9:45 AM Registration
9:35 AM - 10:25 AM Opening Keynotes
10:25 AM - 11:00 AM Break
11:00 AM - 12:15 Talks
12:20 PM - 2:00 PM Lunch - Expo Hall
2:00 PM - 3:00 PM Hands On Workshops (Limited Availability)
3:00 PM - 3:20 PM HOT TAKES
3:25 PM - 3:55 PM Lightning Talks
3:55 PM - 4:45 PM Closing Keynote
4:45 PM - 5:30 PM Booth Crawl Happy Hour (TBD)

Theme we expect to cover

The program is still being shaped, but the event will focus on the practical problems teams face when voice agents move into production.
Expected themes include:

latency and real-time user experience
turn-taking, interruptions, and barge-in
orchestration and dialogue management
evaluation beyond demo scripts
observability, debugging, and production monitoring
human handoff and escalation
compliance, governance, and trust
architecture choices across speech, LLM, and tooling layers

The focus will be on practical lessons from teams building and operating voice systems in the real world, including what worked, what broke, and what changed after launch.

What you’ll leave with

Practical architecture patterns for production voice agents
Better approaches to latency, turn-taking, interruptions, and barge-in
Evaluation ideas that go beyond demo performance and simple benchmarks
Lessons from real deployments, including failures, fixes, and tradeoffs
More concrete ways to think about orchestration, observability, and infrastructure choices
A better understanding of where handoff, compliance, and trust break down in production

Speakers

Speaker announcements are coming soon!

Imagine being in a full room of the best engineers, each one hyper focused on their craft.

Hands-On Workshops
Masterclasses led by experts building voice AI systems at scale.

Bring your laptop.

Leave with frameworks, code, and practical workflows you can implement immediately.

Networking & Community
250+ AI builders, engineers, founders, researchers, and operators. Great food. Meaningful conversations.

20+ Booths

Featuring teams that are actually building production-grade AI agents. Come chat with engineers directly, one-on-one. Ask questions and get the answers you really want.

Interested in a booth? Email dmangabeira@linuxfoundation.org