In the ping pong match of generative AI announcements, it’s once again Salesforce’s turn to make a stir. This past week, the company announced its newest venture, Einstein Studio, which allows users to “BYOM”—bring your own model—to the generative AI sphere. With Einstein Studio, customers can use Salesforce’s custom AI models or their own custom AI models to gain greater insights form their internal, proprietary company data. In turn, businesses will be able to generate smarter, faster insights and content from the data they’re already using via “enterprise-ready” AI.
According to Salesforce, data from Salesforce Data Cloud can be used to train models from throughout the Salesforce ecosystem, including Amazon SageMaker, AWS, Vertex AI (Google Cloud) and others. The company says it can connect and synergize customer data from any source to generate a single customer profile that adapts to customer activity in real time.
Touted benefits include:
● Faster deployment: Einstein Studio doesn’t require businesses to extract, transform, and load (ETL) their data across platforms, which means teams can “point and click” to get their information to the cloud to train their LLMs. This, in turn, means a more productive data/engineering team.
● Generate more content and predictions. Salesforce says Einstein already generates more than 1,000,000,000,000 predictions a week throughout the company’s apps. With Einstein Studio, it will undoubtedly be able to create even more, including content like meeting transcriptions and auto-responses to inquiries, as well as product recommendations, customer segmentation, and personalized pricing. Indeed, it should allow businesses to gain even more insights—and generate even more value—from the data they are already gathering.
● Increase revenue, decrease churn. With greater insights, Einstein Studio will improve the overall customer experience with more personalized and relevant content. This, in turn, will help increase revenue opportunities and decrease churn. In other words, it will enhance the ROI of a company’s investment in AI. Research shows almost 60 percent of companies say they’re a year or more out from implementing AI in their business. Einstein Studio could help make AI more accessible.
Just as importantly, Einstein Studio enables users to determine how their data will be used in training other LLMs, ensuring that enterprise data should be kept safe and secure.
Is Einstein Studio Salesforce’s first entry in the AI race?
Absolutely not. Salesforce launched its first AI solution Einstein in 2016. The strategy then, according to some, is that it would use machine learning, packaged as AI, to differentiate itself across the industry. At that time, however, predictive and generative AI were far less understood than they are now. Einstein’s value, at the time, may have been negligible, but it did create both visibility and intrigue into the power of AI embedded in SaaS—helping pave the way to this year’s AI momentum.
Fast forward to 2023, and it seems like Salesforce is just as committed to finding cutting edge solutions to keep itself relevant while also providing real value to its users. In March, Salesforce announced it would be bringing ChatGPT to its CRM as Einstein GPT. In April, we saw Salesforce begin to integrate its Einstein GPT and Data Cloud into its automation kit called Flow. This allowed users to auto-generate responses using natural language prompts and real-time data sharing to create personalized experiences more quickly. Then, in June, Salesforce announced both Marketing GPT and Commerce GPT to auto generate things like emails, marketing segments, personal shopping experiences and more. And of course, this month, we’re seeing the announcement of Einstein Studio and BYOM. The pace of its offerings in AI have been significant and had to be in order to keep up with commercialization efforts of its competitors.
But is Einstein Studio et al. enough to compete with Microsoft’s Copilot solutions?
The big question now is whether Salesforce’s AI investments set the company up well to compete with Microsoft’s segment of the marketplace. The answer is nuanced because Microsoft’s Copilot and OpenAI partnership is being used across a portfolio that spans well beyond business apps into search, productivity, collaboration and more. Furthermore, Salesforce is leveraging OpenAI as part of its AI stack, which of course benefits Microsoft.
With its current AI related offerings as well as its BYOM announcement, Salesforce’s most direct Microsoft competitor would be Microsoft’s Dynamics 365 offerings including its Sales Copilot, a sales-specific AI assistant. The Sales Copilot tool connects user CRMs with Microsoft 365 and Teams apps to do—well—all the things Salesforce’s solutions aim to do: generate real-time tips for meetings, summaries of customer data, emails, and reports. And, although it’s a Microsoft tool, it can be added to CRMs like Salesforce. Which serves Microsoft well to get parts of the business even if it doesn’t win the CRM part—which is an advantage for Microsoft to play in so many parts of the software suite. But also reduces the competition from all or nothing to harmonious partnerships of heterogenous software, which are often how enterprise software stacks are comprised.
Right now, it’s early to make a call on how Einstein Studio, and Salesforce’s Marketing and Commerce GPT tools will fare in the race to AI and in head-to-head competitive sales. As of now, reviews are scarce, likely because the technology is advancing far faster than enterprise teams can adequately adopt it. Philosophically, I’m positive on flexible AI model deployment strategy as each LLM and smaller language and foundational models bring different attributes. I see such flexibility as a competitive advantage in the long run. Nevertheless, the final story will be told in numbers. But undoubtedly AI is changing the way companies buy and consume software, and I believe that the pace of innovation will only get faster to the benefit of businesses and users everywhere. So, bring on the competition and keep the innovation coming!
Read the full article here
