Amazon Lex Democratizes Alexa-powered Skills

aws-defaultThere’s blockbuster news from AWS re:Invent, the annual gathering of the Amazon Web Services’ community of customers, partners, application specialists and cloud computing experts. The cloud computing giant introduced Amazon Lex, a service that enables its users to build conversational user interfaces (UIs) by employing the same deep learning capabilities that power Alexa, the highly-popular metabot that resides primarily on the 5 million plus Amazon Echos around the world.

Based on testimonials from executives at OhioHealth, Capital One, HubSpot, NASA (all of whom have built “skills” for Alexa on Echo), it is clear that many of the first movers into the metabot space are ready to break out of the rigid Echo form-factors that resides in the friendly confines of the kitchen, family room or bedroom.  They want to bring Alexa’s skills into mobile apps, where most the original bluetooth speakers had been positioned. Even though new form factors like the Echo Dot and Echo Tap have been introduced to expand Alexa’s hearing range within each household, Lex skills will be available through smartphones, tablets, laptop computers, TV remotes, point-of-sale kiosks and, indeed, any device with a microphone.

Lex is in limited release (Amazon calls it “preview mode”). It is only running in AWS’ data center serving the US East, located in Northern Virginia. To encourage development and experimentation it is offered free for those making 10,000 text and 5,000 speech requests each month. After that, it costs $0.75 for every thousand text requests and $4 per thousand speech request  Other terms, conditions, definitions and suggestions developers are provided in this blog post from Jeff Barr, AWS’s chief evangelist.

Lex is designed to make it easy to offer a chatbot on Facebook Messenger immediately. Other messaging platforms and Slack are coming soon. More importantly, the Lex platform has “connectors” to import data from Salesforce, Microsoft Dynamics, Marketo, Zendesk, Quickbooks, Twilio, and HubSpot.  To support generation of proactive alerts, it supports “triggers” based on rules loaded into AWS Lambda event-driven computing service.

There are also hooks into Amazon Cognito, an identity authentication resource that supports simple sign-up, sign-in and data synchronization across multiple apps. This will be key to offering secure, highly personalized services through a Lex-based metabot.

With the introduction of Lex, Amazon has leapfrogged other metabot providers – especially Apple’s Siri and Google Assistant – by opening the metabot ecosystem to a multiplicity of brands, functions and services. It can show the way for more recent entrants, like Samsung Viv, MyWave’s Frank and a handful of other companies trying to popularize the power of a conversational user interface to take command of their digital lives.

The approach is dramatically different from the API-driven approach taken by Facebook, Kik, Slack, Wechat and other popular messaging platforms. They enable tens of thousands of developers to build hundreds of thousands of bots which will measure success based on audience size and frequency of use, much like TV shows and attendant commercials. Metabots, by contrast, have the potential to be true intelligent assistants. They are persistent, ubiquitous and omnipresent conversational interfaces that can be secure, personalized and trusted to carry out queries, commands and transactions on behalf of end-users.

Metabots and basic chatbots are destined to coexist for many years. Still it is Opus Research’s belief that a metabot has significant advantages over the rough-and-tumble of a million or so bots-on-a-messaging-platform. They serve as the highly personalized and secure conversational resource through which an individual can get organized, reach friends, research options, choose things and carry out digital commerce. To each his or her own.

 

 



Categories: Conversational Intelligence, Intelligent Assistants, Intelligent Authentication

Tags: , , ,