Salesforce's Einstein AI makes 1B+ predictions daily. But how much will it cost customers?
- Salesforce's artificial intelligence platform, Einstein, has surpassed the company's expectations and is making more than 1 billion predictions daily, said CEO Mark Benioff at the company's earnings call Wednesday.
- Integrated across Salesforce's clouds and services, the machine learning and deep learning platform has become an integral part of the company's CRM. It is critical in maintaining the company's market lead because competitors have "yet to be able to put in artificial intelligence of a quality, capability and scale that we have it deeply integrated into our application set across the board," according to Benioff.
- Executives dodged around differential pricing of Einstein based on use case complexity, saying it's still in the early days for Einstein and the team will "be thinking very creatively and in an innovative way to sort through" the matter.
The Einstein platform is still relatively young. Salesforce first rolled the technology out in September 2016 across its sales, service, marketing and analytics, community, IoT and app clouds. Salesforce uses the AWS and Google cloud platforms, the latter of which was expanded in November to connect Salesforce customers with Google's Analytics 360 and the G Suite.
Salesforce boasts quite a success story with AI after integrating its capabilities so thoroughly across its platforms to derive so much insight and interaction. But many companies still struggle to implement AI, let alone create a comprehensive business plan around it.
The automated machine learning framework allows Einstein to create AI models tailored to each user at scale. In a Q&A, VP of Data Science Vitaly Gordon said that making Einstein capable of building individualized AI models for every customer required automation. "With 150,000+ customers, the only way to pull off something of this scale without hiring tens of thousands of data scientists is to use AI to automate this process, and that's exactly what my team and I did to build Einstein," said Vitaly. "We built AI to build AI."
Einstein has already been "sitting in" on board meetings for more than a year, reporting what it thinks to Benioff and, on at least one occasion, even pushing back on an employees reports and helping the company solve an unknown problem.
The news about Einstein comes alongside the company breaking records and soaring past preliminary Q4 expectations. Benioff wants the company to double revenue to $20 billion by FY2022, on track with a $60 billion goal by 2034. The company is still secure in its heavy lead in the global CRM market, bolstered by a wide and diverse clientele in which no single client comprises more than 5% of total revenue.
Follow Alex Hickey on Twitter