IBM has touted Watson, its take on artificial intelligence, for at least five years. Salesforce started talking up the technology—at its most basic, a sort of software that teaches itself how to perform tasks—just a few months ago. But the leader in sales software hopes to demonstrate this week that it's ready for this brave new world with smart new online services, some available now, but the majority scheduled to roll out over the next year.
Salesforce chief executive officer Marc Benioff plans to show off these new capabilities, collectively known as Einstein, during the cloud software giant's annual Dreamforce event in San Francisco, along with an application called Quip that will be integrated into its services. Known as a sort of Microsoft Word killer, Quip offers basic document-editing and spreadsheet features. Data in customer records can be pulled directly into Quip documents and spreadsheets; when the record is updated, so too is the related document.
The Quip acquisition, which happened in early August, illustrates the increasing tension between Salesforce (crm) and sometimes-ally Microsoft (msft). The two companies already compete in some core product areas, including sales software. It doesn't help that Salesforce is encouraging European regulators to look into Microsoft's proposed $26.2 billion acquisition of LinkedIn, which it says gives Microsoft an unfair competitive advantage.
Salesforce wants companies to use its online software to manage all of their interactions with customers and customer prospects—from initial contact up to and including all post-sales services. That's a tall order.
The software company has dribbled out tidbits about Einstein for the past month. Another linchpin was revealed yesterday when Salesf0rce disclosed its plan to buy Krux, a marketing software company, for $799 million in cash and stock. Krux's software collects and analyzes data, so that marketing professionals can better promote their organization.
Get Data Sheet, Fortune’s technology newsletter.
Salesforce, which started out in customer relationship management (CRM) applications, has pushed aggressively into marketing services with a series of acquisitions dating back to its $689 million purchase of Buddy Media and the $2.5 billion acquisition of ExactTarget three years ago. It now sells a "cloud" for managing almost every business process—from sales to marketing to service.
Next up is a new service for managing and acting on data related to the Internet of things (IoT). The Salesforce IoT Service cloud, due next year, will meld all that customer sales information with metrics collected from machines and sensors in the field. The idea is to help companies react faster to changing conditions, soaking up input from potentially millions of connected devices.
For more on Salesforce, watch:
"We have all of the CRM data and are bringing it into a single profile to help make the right decisions on handling customers," said Mike Rosenbaum, executive vice president of CRM applications for Salesforce, during a briefing about the company's forthcoming products.
As with all these massive proclamations, it's difficult to suss out what's available in the present versus what will come in the future. But it's clear that Salesforce faces an array of challengers when it comes to dealing with billions of connected "things" or sensors and devices in the field. IBM (ibm), Microsoft, Google (msft), Amazon (amzn) are all attacking this problem in different ways, but artificial intelligence is central to all of those missions. Salesforce rival and partner Oracle (orcl) even touted its own take on AI-driven chatbots at its recent Oracle OpenWorld event.
What Salesforce claims to be doing differently is ensuring that mere mortals performing everyday sales, marketing, and support tasks can benefit from AI. "The key thing about Einstein is accessibility," said Bret Taylor, the chief executive of the Salesforce Quip team and former chief technology officer of Facebook. "In most current cases, you throw AI over the wall and then a legion of programmers are needed to make it work." That, Taylor said, will not be the case with Salesforce Einstein.