AI Integration Breakthrough: Salesforce Einstein Studio Allows Integration of External AI Models

0:00Salesforce has just made a big announcement in AI. And you may notice, you know, Salesforce is not a tech company that has been left behind in this entire AI boom. They've launched Salesforce's Einstein AI and a lot of other really cool AI tools. So today, they're taking it one step further with a new announcement. And we're going to be diving into that on the podcast. Essentially, the headline here is that Salesforce's Einstein Studios lets you bring your own model. And they're going to be starting with Amazon SageMaker. So for context on where Salesforce has been in AI historically, I think they actually first introduced their AI layer of Einstein back in 2016.

0:39But I think the importance of generative AI and data cloud, that's Salesforce's proprietary data lake, was a really key point of discussion at Salesforce World Tour event, which happened in New York City back in May. So I think now the company is kind of taking the next step forward in what they're planning to do. So that's when they have now officially introduced Einstein Studio. So Einstein Studio is Salesforce's kind of latest innovation. It's a management console, which is integrated within the data cloud that essentially allows customers to import existing models without any extract, transform, load, any of those requirements.

1:18So Raul, who is the EV and GM of Unified Data Services and Einstein, he explained, quote, our new feature, Bring Your Own Model, gives customers the ability to incorporate their proprietary data into data cloud to build and train their model. Combining an external model with the Salesforce data and data cloud creates a potent mix. So Einstein Studios essentially is designed for organizations that are, you know, they essentially have sophisticated data teams. They've built models on platforms like SageMaker, but they now want to use those really significant investments they've made, right? They don't want to waste that money and they want to use those in other areas.

2:01And so this solution kind of gives a really smooth way to integrate these models into existing workflows that they already have. And I think this makes a ton of sense. Anytime you've invested heavily into something, especially into AI, you know, developing an AI model somewhere else, and you're like, gee, this would be super useful in my CRM or in other workflows I have. I think this is a really good way to make sure that people are, you know, being able to use their investments to the fullest. So let's talk a little bit about the ETL free advantage. So a standout feature, I think, of the Einstein Studio is its ETL free operation.

2:37So essentially, it eliminates the typical tedious process of extracting, transforming and loading data. This is, I think, a really significant advantage for data teams. And so while it currently supports Amazon SageMaker, Salesforce isn't exclusively going to be working with Amazon. They actually plan to extend their support to Google Vertex, AI, to Databricks, to Snowflake, and to some other platforms soon. So I think this is really awesome. Being a little bit more platform agnostic is definitely going to be really interesting to people. I think especially just because Google Vertex has a lot of really cool tools on it and a lot of really amazing capabilities.

3:17There's definitely enterprises using that building on there. and so I think this is a good move by them to be to be like hey it's not just if you have built something on SageMaker that we have all these other platforms if you built them on the other platforms we also pull them in I think this is going to make Salesforce more powerful and of course this is going to be really beneficial to all the organizations so I think beyond its core function Einstein Studio also offers a big array of kind of predictive models for use cases like customer churn analysis it also facilitates the design of custom custom predictive models so that's you know like predicting product maintenance needs or you know generating product recommendations and they have that all built in so i think the tool also enables the use of language model libraries to generate content such as automated maintenance reminder emails salesforce also kind of aims to reduce model, you know, hallucinations, quote unquote, which of course everyone is concerned about.

4:15And of course that's easier said than done. But I think they're doing this by connecting to an internal graph database, which I think is kind of an interesting approach to this. So this allows for more accurate data driven content and kudos to them for that. So after importing the model, users can integrate into Salesforce's workflows to generate insights or to trigger actions such as email creation. And this really leverages existing work by the data teams that they've already done. And now they're able to just leverage it. So Einstein Studios, with its connectivity to Amazon SageMaker, and the new, you know, bring your own model feature is available in general availability starting today.

4:55And I think that Salesforce's latest developments really underscore its commitment to pushing the boundaries of AI and data management. I think this is going to foster a lot of innovation, which is awesome. And I think that Salesforce really continues to evolve. It's going to be interesting to see how the industry responds. But I think in a really dynamic landscape, you know, with what we're seeing with AI and data management, I think Salesforce's Einstein Studios is an actual potential game changer. I think it really enhances collaboration, customization, and efficiency. And I think this opens up a lot of new avenues for organizations to really capitalize on their data investments that they've already made to not lose those investments, but to be able to carry them through.

5:35And I think this is overall just going to be a really big plus for the Salesforce platform as far as keeping their customers and getting new ones.