Return to site

Along Came Polly (and Lex)

There were more big announcements around voice this week. I’m a feel a bit smug because I wrote about the possibility of this over a month ago. Amazon announced a slew of new AWS services, two of them critical to voice interaction.

Amazon Lex. This is essentially the combination of an ASR and NLU. You feed in audio and Lex spits you back the intents. It’s a subset of what’s available in Alexa Skills Kit to developers and can be used for mobile apps. This could compete with Houndify. More on pricing soon.

Amazon Polly. This is a TTS service — essentially the same voices that were available to IVONA are now rebranded under a new name and a simpler to use API. It’s about time! IVONA since it’s acquisition had been more difficult to reach and it’s great to see this fit in.

While Polly is more about rebranding an acquired technology, Lex is a new offering that directly competes with NLU services on the market. This now offers Amazon a chance to gobble up market share from Google (API.AI, Google Speech Cloud, Parsy McParseface), Nuance Mix, MindMeld and others.

What’s still missing is a dedicated STT service from Amazon. It’s not clear whether they’ll go this route as doing so might allow others to bootstrap their own STT engine or erode the Alexa brand (by completely decoupling a key functionality).

All Posts

Almost done…

We just sent you an email. Please click the link in the email to confirm your subscription!