Introduction

Welcome to Parlant!

Parlant is a structured way to create and manage guided customer-facing AI agents. It lets you build robust agents from scratch; or, in more advanced cases, a robust interaction layer for your prebuilt agents.

Structure is what lets Parlant get better accuracy than most other agentic frameworks. It allows Parlant to run pinpointed, real-time conformance checks that ensure your agents adhere to your instruction. It also gives you specific feedback on where agents may have misinterpreted your instructions, so you can quickly troubleshoot and improve.

Just as importantly, Parlant's way of breaking behavioral guidance down to different elements helps it understand your intentions better. Using this understanding, Parlant actively gives you feedback on your configuration to help you maintain a behavioral configuration for your agents that's coherent and free of confusing contradictions.

And—it's completely open-source, so you're welcome to check out our repo (pun intended) on GitHub: github.com/emcie-co/parlant. (Who's Emcie?)

The Pain Point

Think of an LLM like a stranger with an encyclopedic knowledge of different approaches to every possible situation.

Although incredibly powerful, this combination of versatility and lack of context is also why it so rarely behaves as we'd expect—there are too many options.

This is why, without clear guidelines, an LLM will always draw from its vast but unfiltered set of observations, potentially using tones that are out of touch with the customer and the situation, making irrelevant offers, or just losing focus and going off on tangents.

With Parlant, you can create and serve guided conversational agents to your customers, and—even more importantly—easily improve them over time with feedback from customers and business experts.

The premise behind Parlant is that unguided AI agents are a dead-end. After all, humans are the creative source behind every data point an AI model ever trained on. We should be able to improve their behavior dynamically, inspiring and guiding them to become better performers.

Parlant was designed to let you guide the behavior of customer-facing agents simply, quickly, and reliably—so you can feel confident to iterate on your agents' behavior whenever necessary, teaching them how to best serve your customers.

A focused and guided agent is one that customers could actually use!

So pack your bags and get ready to create some awesome customer-facing agents. Or crappy ones. You've got the controls now and the choice is yours. Let's get started!