Durable Sessions: Why your AI UX keeps breaking and how to fix it
AI products today are being let down not by the models — but by the delivery layer between the agent and the user.
In this session, Fiona Corden, Technical Product Manager, at Ably, breaks down why AI UX is eroding consumer trust, how to spot the delivery-layer problems hiding in your product data, and what the companies getting it right are doing differently. You'll come away knowing how to diagnose whether your AI product has a session layer problem, what durable sessions are, and why they're becoming the standard solution for resilient AI UX at scale.
Topics covered include why AI products break at the transport layer, not the model - and what to do about it. Fiona walks through the five most common agent delivery failures engineers see in production: dropped sessions when network conditions change, silent agents that leave users staring at a spinner, context-blind handoffs where support agents have no history, interrupted returns where users can't pick up an ongoing task from a new device, and trapped users who can't cancel or redirect an agent mid-task. The session includes research findings from 40+ engineering teams shipping AI agents, a live demo of multi-tab sync and agent cancellation via a Vercel AI SDK transport swap, and an architecture explanation of what a durable session actually is and how it differs from durable execution frameworks like Temporal. The Q&A covers Vercel and Cloudflare durable execution gaps, the Temporal integration roadmap, and open source options.
Timestamps
0:00 Introduction & what this webinar covers
2:00 Why AI products are failing at the delivery layer
2:36 Research insights from 40+ engineering teams
3:46 The 5 UX failures users experience today
3:51 Dropped sessions
4:52 Silent agents
5:26 Context-blind handoffs
5:59 The interrupted return
6:42 Trapped users
7:12 What's coming next: multi-user agents & device continuity
8:10 Industry data: consumer trust and AI frustration stats
10:23 Signals in your dashboards
11:40 What great AI UX actually looks like
14:00 Live demo: durable sessions in action
17:46 Where the gap is in your stack
19:34 What is a durable session? Architecture explained
21:09 How durable sessions fix multi-tab, multi-agent delivery
23:06 What this means for your product
25:00 Q&A: durable sessions vs. durable execution (Temporal)
26:15 Q&A: Vercel & Cloudflare durable execution - gaps explained
27:20 Q&A: Temporal integration & open source options
28:45 Closing remarks & next steps
Links
Docs: https://ably.com/docs/ai-transport
AI Transport overview: https://ably.com/ai-transport
GitHub: https://github.com/ably
Ably Discord: https://discord.gg/q89gDHZcBK
Ably on LinkedIn: https://linkedin.com/company/ably-realtime
Ably on X: https://x.com/ablyrealtime
About Ably
Ably is an enterprise-grade realtime messaging platform. We help developers build, ship, and scale realtime features — from chat and notifications to AI agent streaming and live data sync. Millions of messages delivered daily to thousands of companies. Learn more: https://ably.com
#WebSockets #AIAgents #RealtimeAPI #DeveloperTools #AITransport