It’s very likely that you, like many developers today, are exploring how to build AI apps using the hottest tools—and one of them is definitely ElevenLabs.

That’s why we’ve put together a guide that teaches you how to create an app like this—an AI voice assistant we’ve named Pulse:

To try it,

1. Open the demo
2. Press the purple circle and start conversing with the AI 

Here’s the repo: https://github.com/neondatabase-labs/pulse

Breaking down the tech stack

ElevenLabs: The voice behind pulse

At the core of Pulse is ElevenLabs, an AI voice synthesis platform. It’s able to create natural, human-like speech, making it perfect for applications requiring real-time text-to-speech functionality. In our project, ElevenLabs powers the real-time voice responses of Pulse. 

Why we choose ElevenLabs: 

• It has an API-first integration, something essential for our project  
• It’s able to speak in a variety of tones and styles, making it more fun 
• It has subscription-based pricing

Neon: For persistent data storage in Postgres 

Pulse wouldn’t be complete without a solution for managing and storing data—and that’s where Neon comes in. Unlike ephemeral in-memory storage or hardcoded responses, Neon provides a reliable way to persist data, ensuring that Pulse can grow in functionality over time.

Next.js: The web app framework

The front-end and back-end of Pulse are powered by Next.js. We used it for the user interface, API requests, and the real-time updates. 

Why we picked Next.js: 

• Granular control over client side and server-side operations in the application 
• Simple API routes makes it easy to connect with ElevenLabs and Neon

Learn how to build it 

This guide will walk you through how to build an app like Pulse: https://neon.tech/guides/pulse

You can also explore the repository, and ask us any questions in Discord.

ElevenLabs has put together an AI Engineer Starter Pack with discounts for many AI developer tools (including Neon). Claim your deals before they go away!