Whyd now helps companies create their custom voice assistant

Y Combinator-backed startup Whyd is pivoting from hardware to software. The startup had been working on a connected speaker with a voice-control interface specifically designed for music. But a couple of years later, it's clear that subsidized voice assistant devices from Google and Amazon have taken over the market.

Whyd is only keeping its own software platform and partnering with other companies. In other words, if you're working on an app, a website or a skill for the Amazon Echo or Google Home, you can create your own voice assistant to interact with your content.

This way, your users get the same experience across all platforms and you don't have to rely on Amazon's or Google's services.

"We let you integrate with a database of millions of items, create a custom agent and release it," Whyd co-founder and CEO Gilles Poupardin told me. You can think about it as a sort of Algolia for voice queries. Instead of limiting yourself to basic queries ("play my favorite playlist"), you can handle complicated queries ("I want to dance on electronic music").

In particular, Whyd focuses on the cloud infrastructure behind your voice assistant. The company doesn't try to reinvent the wheel and lets you use any speech-to-text SDK. But Whyd can then interpret your query and give you results in little time.

The startup has already worked with 8tracks on its voice assistant. You can now search for music playlists in the mobile app using a voice assistant now. Whyd has developed different models for other verticals. You can imagine a voice assistant for video on demand, e-commerce and other services.

This is what happens between your database and your front end when users interact with their voice: