Chatbots to real-time voice agents: building the next wave of AI at Telnyx

Telnyx is a full-stack platform for real-time conversational AI, bringing together global telephony, AI infrastructure, and flexible APIs so teams can build and scale voice agents on a single platform. Telnyx has over 100 employees in Dublin and AI infrastructure across Europe.

Part 1: Building a real-time voice AI agent
We’ll show you how to build and run a voice AI agent that can talk naturally with users over the phone or web. You’ll see how to connect speech recognition, LLM reasoning, and low-latency voice streaming into a working system, without stitching together a dozen different services.
We’ll cover:

• How voice agents build on the same principles as chatbots
• Key differences: latency, interruptions, and natural speech flow
• Why real-time performance is the hardest (and most exciting) part of voice AI
• Demo: spinning up a voice agent that can answer calls and deal with customer queries.
• Practical tips on scaling from prototype to production

Part 2: Evolving from bots to agentic systems
We’ll also hear from the Telnyx Dublin engineering team, who built an internal support agent to help solve technical customer issues.
As the team lead describes it:

> “We first built chatbots that could quickly answer customer questions, but they were limited to simple back-and-forth responses. Now we’re evolving into agentic systems that can understand what a customer is really trying to achieve, plan the steps, and take action on their behalf.”

It’s not a product, but it’s a real-world look at how chatbot-style automation is being extended into more capable AI agents inside a fast-moving engineering team.