Business software provider Salesforce said on Tuesday it would use artificial intelligence (AI) technology from Amazon's cloud computing unit to improve customer service apps.
Salesforce makes software systems that businesses use to house customer information, like which products or services a customer has purchased and how long they have been a customer. Agents use the information to solve customer issues.
Salesforce said it will use technology from Amazon Web Services to turn the customer's spoken words into text, where it can be translated into different languages or analysed to determine whether the customer is angry or satisfied, all in real time.
From there, Salesforce's own software can read the text and make suggestions using data already in the business' Salesforce-based systems, such as recommending answers to the customer's questions.
Bret Taylor, Salesforce's president and chief product officer, said the practical effect of the software could be as simple as not having to put a customer on hold.
For instance, he said, when a customer calls an insurance company to ask the size of a policy's deductible, instead of an agent putting a customer on hold to look up the data, the software can fish it out of the system automatically.
"It really shortens the time to resolution on these calls, which is really great for both sides of that interaction," Taylor said in an interview. "For people who are in a contact center, improving the efficiency of all those calls is a huge cost (savings). And for you as a consumer, it means you get off that call more quickly, you solve your problem more quickly."
Salesforce plans to unveil the software at a user conference in San Francisco on Tuesday, which draws tens of thousands of attendees to the city.
The partnership brings Salesforce and Amazon into more direct competition with call center technology providers such as RingCentral. RingCentral has built out cloud technology for call centers and recently executed a deal to supply Avaya, which was evaluating sale or merger options after struggling to build out similar technology on its own.
(Reporting by Stephen Nellis in San Francisco; Editing by Christopher Cushing)