Listen to this story
Recently, at the first-ever developer conference – DevDay 2023, OpenAI chief Sam Altman revealed that the company has about 2 million developers building on their API for a wide variety of use cases, and over 92% of Fortune 500 companies build on their products.
Salesforce’s Einstein GPT is one such model built on top of one of these APIs, which it introduced in March for CRM applications. Salesforce CEO Marc Benioff believes that it is top of its game, and said that it is already the leading global artificial intelligence platform.
“The whole company is pivoting around this next level of AI, and Salesforce now really is betting on the future of AI plus, data plus, CRM,” said Sridhar H Hariharasubramanian, senior director of solution engineering at Salesforce, in an interaction with AIM.
Subscribe to our Newsletter
Join our editors every weekday evening as they steer you through the most significant news of the day, introduce you to fresh perspectives, and provide unexpected moments of joy
He said, its platform Einstein AI and Einstein Copilot (built on Einstein GPT) generates over 100 billion predictions daily and has found real-world customers for the same.
“We believe that this combination of applying artificial intelligence on top of enterprise data, and combining that with the insights that you are going to draw from our CRM platform, is going to really deliver a lot of benefits in terms of delivering better customer experience,” he added, saying that all of this would bring in a lot of improvement in operational efficiency, productivity and growth, for these organizations.
Salesforce’s vast pool of customers also poses as brownie points for them. With an unprecedented market share of 32.94% in CRM, it has over 149,665 customers. While the three top industries that use Salesforce CRM for salesforce automation are marketing (2,485), technology (2,338), and education (2,130), it also has over 3472 customers across AI and ML
Unleashing Enterprise Innovation
“We have generative AI capabilities in sales cloud, in service cloud, marketing cloud, commerce cloud as well as our data cloud product. We also have our own LLM called Code Gen, which is actually a purpose-built or a custom-built model, which will help companies who are large client of Salesforce, to write code and applications without actually doing a lot of coding,” he said.
“So you can simply in a typical GPT kind of fashion provide instructions in English with some logic in it, and the code will automatically get generated. And, of course, a human can review the code, it’ll put in the syntax if you put in remarks,” he added. Further adding “We have these generic AI capabilities across pretty much all of our key offerings.
The Einstein Copilot, powered by Einstein GPT and CRM insights, can also enhance productivity and efficiency in various business tasks. It assists users, such as salespeople, in generating emails or call summaries by providing contextual and customised content.
For instance, in sales, Copilot can help draft prospecting emails tailored to individual clients, combining user instructions with CRM data for personalized communication. The tool’s applications extend to service-related tasks, summarizing knowledge articles or generating responses for client inquiries.
Notably, the generated content undergoes human review before finalization and is intended to save time and improve the relevance of communications. Sridhar highlighted another key feature—Salesforce’s new Einstein 1 platform, introduced at Dreamforce.
The platform unifies metadata frameworks across various applications such as sales, service, marketing, commerce, Tableau, and industry clouds. This integration enhances the seamless exchange of information across these apps, providing a more cohesive experience. The platform’s unified architecture reduces the reliance on data extraction and loading processes, said Sridhar.
For instance, Einstein Copilot is integrated into Tableau, allowing users to request key highlights from complex reports using natural language, enhancing accessibility for business users.
Building Trust in the Era of Generative AI
Salesforce said that it is committed to addressing concerns related to trust and data privacy.
Sridhar clarified that while Salesforce partners with AWS for cloud services, all of Salesforce’s capabilities will be available on Salesforce’s native platform, Hyperforce—which ensures data privacy and regulatory compliance, making it the ideal choice for Indian customers.
Salesforce’s differentiation also lies in the Einstein Trust Layer, addressing enterprise concerns about data privacy and security. This layer incorporates dynamic grounding, contextualising prompts by drawing from correlated data before reaching foundational models to prevent inaccuracies.
He further said that data masking ensures no personally identifiable information (PII) is shared externally. The zero retention architecture guarantees that OpenAI, the LLM provider, won’t retain customer data, addressing worries about data use for training algorithms. Toxicity detection assesses response appropriateness, addressing biases. “An audit trail ensures transparency and trust, setting Salesforce apart in the CRM space,” he added.
Indian Market Thrives
Sridhar added, “India we are seeing a lot of demand for it.”
The company also grew at an unprecedented 24% YoY in Q2FY24 in the APAC region.
He said that the reception of Einstein Copilot and other new capabilities has been encouraging and clients are eager to leverage generative AI to drive growth, increase productivity, and deliver better experiences, among other things.
“I think the feedback has been very encouraging for at least the last two, or three months since some of these announcements started coming. We’ve had fairly serious conversations because Indian customers are obviously aware of these developments”