Overview

Develop conversational AI applications with NLX

AI applications are the core of how you build, run, and deliver conversational experiences with NLX. Whether you’re creating a simple chat assistant, a hands-free Voice+™ experience, or a fully autonomous agent capable of executing complex workflows, every solution starts with an application.

This chapter introduces the different types of AI applications you can build, how they run, and how they’re deployed across channels and environments.

What are AI applications?

An AI application is the deployable package that brings your conversational experience to life. Each application bundles your flows, knowledge bases, language settings, integrations, and delivery channels into a single unit that you can build, deploy, and iterate on.

Once deployed, your application becomes the backend runtime that powers conversations wherever your users interact, including web, mobile, voice, IVR, and more.

AI application in the NLX platform

What you can build

NLX supports multiple types of AI applications, each optimized for different user interactions and levels of automation.

Core

The foundation of NLX's conversational automation and the most customizable application type

Voice+

A hands-free voice experience that synchronizes spoken commands with on-page navigation, form filling, and contextual actions

Agentic

A powerful application type that lets your AI reason, plan, and execute multi-step tasks using LLMs, APIs, tools, and custom workflows

Async Worker

Fully autonomous background agents capable of running tasks independently, such as researching the web or generating content

How you can deploy

Deploying an application means packaging your flows, configuration, and resources into a runnable version and pushing it to the channels where it will operate.

Each NLX application can be deployed in multiple ways, across channels, websites, hosted NLX Touchpoint apps, or as an MCP server, depending on where and how you want users (or other AI clients) to interact with it.

Last updated