PolyAI

Starting today, we're opening our Agentic Dialog Platform to every enterprise builder. Our dialog agents have resolved 1 billion+ customer conversations for clients like FedEx, Unicredit, PG&E, Marriott, Foot Locker, and many more. These aren't easy conversations. They solve problems like: > A patient booking medical transport who needs insurance verified on the spot. > A homeowner calling their utility company about a gas leak. > A cardholder figuring out why their must-have purchase was declined. Standard conversational AI was never built for this. It was designed for chat, adapted for voice later. It generates responses, but can't do what dialog requires: hold context under pressure, navigate ambiguity in real time, and actually resolve problems. So we built a better model. Our proprietary model Raven was built from the ground up specifically for dialog. Agent harness in the weights, not bolted on through prompts that drift under pressure. And in our platform, you can deploy Raven as your default or bring in GPT-5, Claude, Gemini, whatever model fits your use case or regulatory requirement. Now that the Agentic Dialog Platform is open, any team can create, test, and deploy dialog agents on the same model and infrastructure the world’s top brands trust on their hardest days. This opens up the pool of builders across your entire enterprise. The person who knows customers best, who runs operations, who owns the customer journey: they're all builders now. Two ways to build: > Poly Agent Builder: Describe your use case in natural language, and it configures your agent, knowledge base, and conversation flows automatically. Production-ready in ten minutes. > Agent Development Kit (ADK): Developers use this to build dialog agents the same way they build everything else. Use your own IDE, a coding assistant like Claude, version with Git, deploy from your terminal. Get started now: studio.poly.ai

What does AI actually change for large engineering teams? In this episode of Deep Learning with PolyAI, Nikola Mrkšić sits down with Helen Greul, PolyAI's recently appointed SVP of Engineering, to explore how engineering teams are moving toward smaller, more nimble units, what that looks like from inside a fast-moving AI company, and what it means for leaders trying to build and manage technical teams today. Watch the full episode: youtube.com/watch?v=E6rKIU…

Deploying an agent shouldn't feel like a risk event. Manually testing agents usually takes hours and still only covers the scenarios you thought of. A customer says, "Please don't send what I ordered last week, I don't want it anymore." Your agent was trained to recognize "cancel my order" and defaults to fallback, so a routine request becomes a frustrated escalation. You need a way to know your agent handles the scenarios you didn't think to test. We've built testing inside Agent Studio with three capabilities working together: 1. AI test generation Studio Assistant builds test scenarios directly from your agent's flows, knowledge base, and tools. Broad coverage from day one, without manual scenario-writing. 2. Simulation testing Run hundreds of realistic conversations against your agent at once. Define scenarios in plain English, let an LLM judge evaluate every outcome, and fix issues before they reach a customer. 3. A/B testing Run two versions live. Split traffic, compare containment rate, CSAT, and handle time, and promote the winner once the data backs it up. When every change is validated before it reaches a customer, your team moves with confidence. Learn more about PolyAI's testing capabilities: poly.ai/blog/testing-a…

@FastCompany just ran a feature on some of the most important companies building AI right now, and PolyAI features for our work in voice. The piece looks at the infrastructure, governance, and application layers that make AI commercially viable at scale. For us, that means turning billions of customer conversations that used to go unheard into real resolutions, the kind Forrester found deliver 391% ROI and $10.3M in savings over three years. Read the full piece here: fastcompany.com/91564497/you-k….

How do you keep a team fast and aligned at the same time? On the latest Deep Learning with PolyAI, Nikola Mrkšić sits down with PolyAI's SVP of Engineering, Helen Greul, to discuss: → Why engineering is moving to smaller, nimble units connected by shared context → How autonomy and alignment can live together → Why great engineering leadership is evolving from T-shaped to W-shaped and star-shaped → Why the next generation of great builders won't necessarily come from a traditional engineering background Check out the full episode: youtube.com/watch?v=E6rKIU…

A financial services company handed its bereavement calls to a voice AI agent. It scored higher on customer satisfaction than the human team. The calls teams most want to protect from automation are often the ones it handles best. PolyAI's CEO, Nikola Mrkšić, joined Dominic Monkhouse on Scale to Win to talk through what that means for businesses still deciding where AI belongs. Watch here: bit.ly/4w3oD1d

Most AI agents running in contact centers today are built on general-purpose models with defaults designed for long-form text and screen-based interactions. On a live customer call, that means responses that are too long, phrased for a reader, and delivered at a pace that makes callers wonder if anyone's listening. Raven is how we solved it. A model built specifically for customer service, trained on millions of real conversations, and designed for how people actually talk. In our upcoming session, Matt Henderson and Jak Katterfield will walk through why we train Raven in-house, how we do it, and what it delivers on a live call. Join us July 1st at 12 PM ET: poly.ai/webinars/not-a…

Hawksmoor has spent 20 years building a hospitality culture that takes pride in every guest interaction. Especially the first one, which, for many guests, is a phone call. At peak times, they were missing as many as 42,000 calls a year, with no visibility into what those callers needed. Now PolyAI answers every call across all 14 sites, around the clock: → 80,000 covers booked → Missed call rate reduced from 20% to zero → 90% guest satisfaction score For a brand built on the quality of every touchpoint, the phone now delivers to the same standard as everything else they do. Hear from Alex Grace-Smith, Hawksmoor's Group Head of Reservations, and read more about Hawksmoor's story: poly.ai/en/customers/H…

Made in London with AWS: Meet the founders. London has long been a gateway to global commerce. Today, a new generation of founders is continuing that legacy, building and scaling ideas and innovation. We went behind the scenes with VC firm @BessemerVP and trailblazing startups Apoha, @polyaivoice, @ValyuOfficial and Zego to explore what makes London a launchpad for world-changing ideas, and why it continues to draw global talent and investment.

Learn more about Adopt 100 here: deloitte.com/uk/en/about/pr…

Through Adopt 100, it’s become easier than ever for Europe's largest enterprises and private equity portfolios to adopt our platform leveraging NVIDIA’s technical infrastructure and Deloitte's transformation expertise.

Europe has no shortage of AI innovation. Getting AI into production is where things stall. Today, we're proud to announce we’ve joined Adopt 100: a new program from @nvidia and @Deloitte designed to bring AI into core enterprise operations.

When you handle over a million calls a month for members getting to dialysis and oncology appointments, every single one has to be answered. SafeRide Health's call volume doubled year over year, and hiring enough people to be there at all times wasn't possible. They needed a way to be there for members in the moments that mattered most. That's why they chose PolyAI. Together, we created an agent with a voice that their members could trust, with the reliability to perform at scale, and the accuracy to get every interaction right. Steven, their PolyAI agent, now handles over a million calls a month, ensuring every member gets through. They've saved 47,000 agent hours on authentication alone. As Ben Salter, Chief Product Officer and Co-Founder at SafeRide Health, says: "If you want a 99% on-time fulfillment rate, a 4.8-star member rating, and 97% on-time performance, you need AI automation. It's the only way you maintain member experience and costs." Read the full story: poly.ai/en/customers/S…

🇨🇦 PolyAI is coming to Toronto! Toronto has one of the deepest concentrations of AI talent anywhere in the world, and our North American customers are growing fast. Being on the ground means we can move with them and recruit the teams who will define the next phase of our platform. Our team there will be focused on agent design, deployment engineering, and business development. Read about where we're headed next: poly.ai/blog/PolyAI-ma…

Our Agentic Dialog Platform is now open to every enterprise builder. On the latest Deep Learning, PolyAI's Senior Forward Deployed AI Engineer, Ruari Phipps, and Director of Product Management, Frank Ferro, show what that looks like in practice. Using the Agent Development Kit (ADK), they go from a blank project and a one-page spec to a fully working agent that handles knowledge base Q&A and a multi-step booking flow, plus is ready to deploy across voice, chat, and SMS. However you prefer to build, there's a path for you. Describe what you need in natural language with our Agent Builder, or go code-first with the ADK: your IDE, your Git workflow, your terminal. Build your agent today at studio.poly.ai Watch the full episode: youtube.com/watch?v=x3TDve…

@bally_kehal We're excited to hear what you build with it: studio.poly.ai

@polyaivoice Voice AI has been demos for too long. This is the first platform I've seen where the model is actually built for dialog, not adapted for it. Big launch.

Starting today, we're opening our Agentic Dialog Platform to every enterprise builder. Our dialog agents have resolved 1 billion+ customer conversations for clients like FedEx, Unicredit, PG&E, Marriott, Foot Locker, and many more. These aren't easy conversations. They solve problems like: > A patient booking medical transport who needs insurance verified on the spot. > A homeowner calling their utility company about a gas leak. > A cardholder figuring out why their must-have purchase was declined. Standard conversational AI was never built for this. It was designed for chat, adapted for voice later. It generates responses, but can't do what dialog requires: hold context under pressure, navigate ambiguity in real time, and actually resolve problems. So we built a better model. Our proprietary model Raven was built from the ground up specifically for dialog. Agent harness in the weights, not bolted on through prompts that drift under pressure. And in our platform, you can deploy Raven as your default or bring in GPT-5, Claude, Gemini, whatever model fits your use case or regulatory requirement. Now that the Agentic Dialog Platform is open, any team can create, test, and deploy dialog agents on the same model and infrastructure the world’s top brands trust on their hardest days. This opens up the pool of builders across your entire enterprise. The person who knows customers best, who runs operations, who owns the customer journey: they're all builders now. Two ways to build: > Poly Agent Builder: Describe your use case in natural language, and it configures your agent, knowledge base, and conversation flows automatically. Production-ready in ten minutes. > Agent Development Kit (ADK): Developers use this to build dialog agents the same way they build everything else. Use your own IDE, a coding assistant like Claude, version with Git, deploy from your terminal. Get started now: studio.poly.ai

@saidul_dev That's exactly the distinction that matters. Generating a response and actually resolving a problem are very different things.

@polyaivoice The interesting shift here is moving from simple conversational AI to systems built for real-time problem resolution.

@AhsanJaveriya Those conversations are exactly what shaped how we built Raven.

@polyaivoice The shift from “chat AI” to “dialog AI” is the next big leap. You nailed the problem statement perfectly. The patient transport and utility examples hit home, those are nightmare scenarios for generic LLMs.

@trikcode A lot of hard work went into getting here! We're excited for you to try it: studio.poly.ai

@polyaivoice This team is doing amazing work and pushing boundaries

@JafarNajafov Exactly! The people who understand the problem best should have access to the solution 🙌

@polyaivoice The real unlock here is giving the “customer expert” builder access. Most enterprises don’t lack AI tools. They lack a way for the people closest to the customer to turn their knowledge into working systems. That’s where this gets interesting.