PolyAI

Deflection rate has become the metric leaders use to justify automation investments, but a high deflection rate doesn't mean customers are getting what they need. In an interview with Customer Experience Magazine (CXM), PolyAI CEO and Co-Founder Nikola Mrkšić explains why resolution is the metric contact centers should be tracking. Here's the case he makes: → Deflection only confirms a customer was kept away from an agent, while resolution measures whether their problem was actually solved → "Digital deflection," where customers get redirected, logged as contained, then dropped, registers as a win even when nothing was resolved → Unresolved issues come back as repeat contacts, channel switching, and churn → PolyAI's data shows that when relative resolution approaches 100%, customer satisfaction scores reach five out of five Read the full interview with CXM: https://lnkd.in/eJ9xBD-9

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: https://lnkd.in/eHvrPbuh

We just opened our Agentic Dialog Platform to every enterprise builder in the world. For the first time, anyone can build, test, and deploy dialog agents in minutes on the same platform Marriott, UniCredit, Foot Locker, PG&E, and hundreds of the world's most demanding enterprises trust for their hardest conversations. Dialog is our new interface for building, which means you've got a whole new pool of potential builders across your entire business: the person who knows customers best, the person who runs operations, the person who owns the customer journey, anyone with an idea and an email address. Here's what you can access, starting right now: → Poly Agent Builder: put natural language in, get a production-ready agent out. Describe your business needs, and we'll configure the agent, knowledge base, and conversation tracks and guardrails automatically in minutes, drawing on knowledge distilled from hundreds of enterprise deployments. From there: test it live, analyze call data, and refine through an ongoing dialog with the platform until every conversation is exactly right. → Agent Development Kit (ADK): With self-serve API keys, native integrations, and full CLI support, the ADK lets developers ship dialog agents the same way they ship everything else. Build in your own IDE, use a coding assistant like Claude, version it with Git, and deploy from your terminal. → Shareable testing: Every agent ships with a zero-setup, shareable test environment that you can interact with live across channels. Get stakeholder buy-in and validate behavior before anything goes to production. Whether you're a product manager, operations lead, CX pro, or developer, you can start building today at https://lnkd.in/gsvBuPMf 

PolyAI's Co-Founder and CTO, Shawn Wen, took the stage twice at CCW UK this week. His keynote covered how enterprises are moving from software that supports work to agentic systems that do the work. As AI agents understand context, take action, and connect intelligence across every customer interaction, legacy tools built for human workflows start to break down. The defining challenge throughout it all is maintaining customer trust. He also joined Rory Sutherland for a fireside chat on the psychology of customer experience. Some takeaways from the session include: → Organizations invest heavily in functional improvements, but they don't address the fact that the entire customer interaction is structured around the business, not the caller. The customer walks away feeling like they did the work. → Business logic says faster is always better. The behavioral reality is that many callers, elderly customers, first-time users, people calling about something stressful, need breathing room, not speed. → The model can be brilliant, but if the interaction is poorly designed, it still feels like "talking to a vending machine." The design layer on top, the thing that turns a capable system into one that feels engaging, is where most organizations underinvest. Learn how we make design decisions at PolyAI to create conversations that actually work for customers: https://lnkd.in/eY8Zr9dK

Imagine a customer calls in Spanish. Your AI agent picks up, and the conversation starts well. Then, somewhere around turn eight, the agent loses track of the language it's supposed to be speaking and reverts to English. This is what happens when models reason in their most comfortable language and translate on the way out. Language adherence is built into Raven, PolyAI's model purpose-built for customer conversations, so you can prompt it in one language and have it respond in any other language consistently across 23 languages. Matt Henderson explains why multilingual reliability is one of the things that sets a purpose-built model apart. Learn more about the importance of specialized AI models: https://lnkd.in/e6ajpnHX

When you're handling millions of customer conversations, the insights are in there. The hard part has always been getting to them. That's what Deep Sampling in PolyAI's Smart Analyst is built to do. Instead of manually reviewing transcripts or working from anecdotal feedback, you can prompt it with questions about your data, analyze hundreds of conversations, and surface what's actually happening. PolyAI's Greg Janota shows you how. Find out more about how to gather and act on your insights at scale: https://lnkd.in/e3r3e2gu

PolyAI is growing fast, and we're bringing in a bench of incredible leaders to keep building. Helen Greul joins as SVP, Product Engineering. Helen brings more than 15 years of experience building scalable systems that enterprise customers can rely on, most recently leading engineering at Multiverse and before that at Spotify. MORGAN Sage N. joins as Chief Marketing Officer. A four-time CMO with senior roles at Microsoft, Zuora, Cohere, and Dialpad, Morgan knows how to build bold, competitive marketing in crowded enterprise markets. Paul Asoyan joins as VP, Strategic Alliances and Business Development. One of the founding members of Google Cloud's partner ecosystem, Paul most recently led partnerships at Gong. He'll be building and scaling ours. Danielle Waknine joins our Canadian operations to lead GTM expansion into new verticals. She comes from Shopify, where she scaled the B2B vertical from zero to $100M in under a year. Please join us in welcoming all four on board! Read more about the hires, our new Toronto hub, and where we're headed next: https://lnkd.in/e9wx8Bix

We're heading to Customer Contact Week UK next week. Shawn Wen will join Rory Sutherland of Ogilvy UK for a fireside chat on the psychology of customer experience. The conversation brings together a technical view on building AI for real customer interactions and decades of behavioral science around how people actually make decisions. If you're at the event, come and find us at the Mainstage at Novotel London West, Tuesday, May 12th, 9:05 AM. https://lnkd.in/es54YQfN

Enterprises are moving from software that supports work to agentic systems that do the work. As AI agents understand context, take action, and connect intelligence across every customer interaction, legacy tools built for human workflows break down. The defining challenge is maintaining customer trust. Shawn Wen, PolyAI's CTO and Co-Founder, shares real case studies on building and scaling AI-driven CX that earns and sustains that trust at Customer Contact Week next Tuesday. See you there: https://lnkd.in/es54YQfN

The guest experience starts long before a customer walks through a restaurant's doors. But the staff who shape that first impression are already stretched. More restaurants are using AI to take the pressure off, so teams can focus on the people in the room. The Financial Times looks at how hospitality businesses are putting AI to work across front and back of house, including how OpenTable is using PolyAI to handle reservations: https://bit.ly/3QZ3ENN