Meet Wren: the agent that looks after your agents. Your dialog agents handle thousands of customer conversations a day, and keeping watch over all of them to make sure everything is working as well as possible is a lot for any team to stay on top of. Overnight, it reviews conversations, catches bugs like a validation rule rejecting valid customers, and has a fix ready before you log on. Live now. Read all about it here: bit.ly/4iNF5ze

“Hello? Is anyone there?” 🤔 That’s what callers say when a voice agent takes a beat too long. AI can write production code, but many agents still can’t tell when to speak on a live call. Dialog-RSN-1 hears the caller’s audio and handles the whole turn in a single pass, so callers hear a reply in under 1 second. Independent benchmarking by Coval found it the strongest realtime model they’ve tested on production-style calls. @pmckee, @Addlesee_AI, and @matthen2 take you inside the model in Part 1 of Let’s Dialog: Building the Next Voice Agents. Grab your spot: bit.ly/4dVUc6B

Guten Tag from #GTCBerlin! We’re excited to be featured in @nvidia's healthcare news for our multilingual deployment with @corilusnews, built on NVIDIA Agent Toolkit. Their network spans 70,000+ healthcare pros in Belgium and France. Here this week? Say hello. Read the full announcement here: nvidianews.nvidia.com/news/h…

We assume the sensitive calls are the ones that need a human. Ian Jacobs of @opusresearch says the data often shows the opposite. When a conversation carries social friction, taking the human out of it can push CSAT and other metrics up. His example is collections: it's embarrassing to set up a payment plan with a person on the line, so people are more likely to go through with it when they're talking to AI. One customer saw promise-to-pay rise once conversational AI took over. Where else does automation beat a human? Watch the full conversation here: bit.ly/3SvI6ts

We're in Berlin for NVIDIA GTC this week. Stop by booth #7045 and put our dialog agents to the test yourself. Interested in setting up time with our team? Learn more here: bit.ly/4y1UXSR

Dialog-RSN-1 is now GA, and anyone can get an API key and build on it today. Comparing real-time voice models? Start with the numbers. Median first-token latency is under 300ms, about the length of a blink. It has the highest quality, audio-understanding, and safety scores of any real-time-capable model, and it’s efficient enough to scale to every conversation you have. It’s audio-native, so turn-taking, speech recognition, function calling, and the response all happen in a single pass, with a separate voice output you control. Get started in three steps: 1. Pick your path: our Agent Development Kit (ADK), AWS SageMaker, or the public API, wherever your stack already lives. 2. Create your account and grab your API key. 3. Clone an example, add your key, and talk to it in your browser. Start now here: studio.poly.ai/start-buildin…

Two big wins for PolyAI. Best AI Tool at the 2026 @Digiday Technology Awards, and a place on the @FinancialTimes first ranking of the UK’s fastest-growing companies. 🏆 Both trace back to hundreds of enterprises running frontier dialog on our platform across 75 languages and 25 countries. This year we gave them Dialog-RSN-1, the world’s fastest reasoning model for real-time complex dialog, and Wren, which keeps improving their dialog agents so performance compounds over time. Thank you to our customers, the driving force behind all of it. Read more about the wins here: poly.ai/blog/digiday-best-ai…

Next week, PolyAI is at Frost & Sullivan Contact Center West. 🌵 Come find us on the show floor and catch our session with AdaptHealth: “How AdaptHealth Unlocked AI Agent Leverage: From First Use Case to Enterprise Scale.” Bill Lewis, Senior Director, Communications Technology at AdaptHealth, and PolyAI’s Jose Carlos Berlioz, Senior Solutions Consultant, will cover the governance, team structure, and use-case prioritization that turn an early win into something the whole business builds on.

When Wren flags a problem or opportunity with your dialog agent, it doesn't just tell you that it’s time to make a change. It shows you the actual calls, drafts the change itself, then simulates the fix and tests it before you approve anything. And it keeps a record, so the same problem can't quietly happen again. See how Wren will gain your trust in the full episode: bit.ly/4cNE1Yq

One week until NVIDIA GTC Berlin. Booth #7045 slots are filling up. If you're scoping a voice AI project, let's talk about what it actually takes to get from prototype to production, and where most attempts stall out. Grab some time with our team here: bit.ly/4jYlblw

“The next challenge is to make reasoning very fast, so that the models can fetch answers quickly and the conversation feels natural.” Our CTO Shawn Wen on why voice AI hasn’t had its ChatGPT moment yet, via @TechCrunch. Fast reasoning for dialog is what Dialog-RSN-1 delivers: under 300 ms responses in live calls. Read the full story here: bit.ly/4hJojR3

One week until Gartner IT Symposium/Xpo. Our CEO Nikola Mrkšić takes the stage Oct 21st to separate agentic AI that works from what just demos well. Lead CX or contact centers? Request an invite for our exclusive dinner on Oct 19th. Want to meet our team in Orlando? Learn more here: bit.ly/4hLKN3y

One week, two CCW events, and a lot of candid conversations about putting AI in front of customers. At CCW Europe, our CEO @nikola_mrksic moderated the mainstage panel “Winning Over the Room: How CX Leaders Are Building the Case for Customer-Facing AI” with CX leaders from NatWest and UniCredit, who shared how they built buy-in, worked through governance, and took AI from pilot to production. Then it was on to CCW Nashville for booth conversations about what dialog agents can do on your customer calls, and a happy hour at Urban Cowboy. Thank you to everyone who came to the panel, stopped by our booth, or pulled up a chair.

Please help us welcome Sunil Shah as Chief Financial Officer at PolyAI. He brings 20 years of experience growing software businesses, and a view of growth from both sides of the table as an investor and an operator. Read more about this exciting addition here: bit.ly/3UgfO6X

Build a working dialog agent in 90 minutes. Join us at PolyAI HQ in London on October 14, 2026 for AI After Hours, co-hosted with AWS. Bring a laptop, we cover platform access, food and drinks (tacos and margs included!). 🌮 Only a few spots left. Sign up here: bit.ly/4yCvpN9

Dialog-RSN-1, the fastest reasoning model for real-time complex dialog, is now on AWS SageMaker. Deploy it as an endpoint in your own AWS environment, no separate speech-to-text, sub-300ms responses, and results like an 11% containment lift and 37% cut in latency. Read more about this exciting partnership here: bit.ly/4ze53kl

Our CTO Shawn Wen sat down with Tim Scarfe of Machine Learning Street Talk to talk about why voice is still the hardest problem in enterprise AI. Text gave models all the time they needed. Voice gives them a fraction of a second to read your pauses, your mood, your tone correctly. Read the full breakdown of the episode here: bit.ly/4APgqAL

Some of the best conversations happen before 9 AM. ☕ Earlier this week, CX and ops leaders joined us for a private breakfast in London on where voice AI delivers in the contact center, and where it still falls short. Thanks to everyone who joined us. We're already looking forward to the next one.