SageMaker has long been AWS’s fully managed platform for building, training and deploying machine learning and generative AI models. Over time, however, an ecosystem of applications has sprung up around AI and ML models for performing tasks like managing experiments, evaluating model quality and security. Those always lived outside of SageMaker and had to be managed separately, but now, at its annual re:Invent conference in Las Vegas, AWS is launching a first curated set of AI apps that will be fully managed and secured by the SageMaker team.
Some of the early partners who are making their applications available in SageMaker are Comet, Deepchecks, Fiddler, and Lakera Guard.
“One thing that we’ve heard from customers is that, as they’re building out end-to-end AI solutions and workflows, oftentimes they want to use a purpose-built tool — a third-party tool — which they really like, and they want it to work really well with their SageMaker development environment. But today, what they have to do is they have to spend time and effort in integrating those third-party tools with the rest of the Sagemaker system,” Ankur Mehrotra, the director and GM for SageMaker at AWS, told me.
Mehrotra also noted that many companies worry about the security implications of using a third-party tool since they don’t want their data to be shared with these additional services. With these new AI apps inside of SageMaker, AWS says, all of a company’s data will stay within the SageMaker environment.
