AWS aims to bring machine learning, natural language processing to call center


Asha Barbaschow

By Asha Barbaschow

| December 3, 2019 — 19:14 GMT (19:14 GMT)

| Topic: Cloud


Amazon Web Services (AWS) has packaged up another in-house Amazon capability and made it available to customers, having announced Contact Lens for Amazon Connect on Tuesday.

CEO Andy Jassy has touted that Contact Lens for Amazon Connect — the company’s omnichannel cloud contact centre service — will stitch together new abstractions for machine learning so AWS customers can have an easy to consume function.

“Amazon Connect is one of the fastest growing services in the history of AWS … off to a blazing start,” he said during the day one keynote of AWS re:Invent in Las Vegas. “Using the same customer service technology Amazon has used … it’s really easy to use.”

According to Jassy, Connect is the first call centre in the cloud with machine learning in mind.

“People could have easily built it themselves on top of Lex, but they love that we do that,” he said.

The service can be activated through a single click in Connect; it automatically transcribes and analyses customer calls, and also has the capability to transcribe previously recorded calls.

Most importantly, Jassy said, it allows users to determine the sentiment of the call, pick up on long periods of silence, and times when an agent and customer are talking over the top of each other.

After already announcing a fraud detection service that packages up an existing in-house product, Jassy said another area his company can offer up pre-existing Amazon capabilities is in code.  

“If there’s a problem with the code, these other steps don’t matter — you’re going to have a bad customer experience,” he said, discussing customer service

The CEO unveiled Amazon CodeGuru, which is a new machine learning service to automate code reviews and identify the “most expensive line of code”.

“Millions of code reviews we’ve done at Amazon … 10,000 most popular open source projects,” Jassy said, touching on why he thinks AWS has the best capabilities to help with code review.

CodeGuru will see customers write and commit code, then add CodeGuru to pull requests — the service is currently integrated with GitHub and CodeCommit. Jassy said more are on their way.

CodeGuru then reviews the code based on reviews from Amazon and those open source projects Jassy touched on.

The service then comments on pull request.

“It provides an assessment of your code,” Jassy said.

“When we shared CodeGuru with a few customers, just this first piece was a game-changer.”

Asha Barbaschow travelled to re:Invent as a guest of AWS.      

Related Topics:


Digital Transformation

Data Centers




Asha Barbaschow

By Asha Barbaschow

| December 3, 2019 — 19:14 GMT (19:14 GMT)

| Topic: Cloud