Domino Data Lab on Tuesday introduced new capabilities within its Enterprise AI platform.
The new features will help enterprises accelerate the production of generative AI and traditional AI responsibly, according to the AI and MLOps vendor.
To help with acceleration, Domino introduced the AI Project Hub, a marketplace, or app store, built into the Domino Enterprise AI platform. It is available now.
The hub grants enterprises access to popular AI applications. It provides tools, workflows and models that users can customize. Enterprises using the hub can access models and tools from Domino Data Lab's partners, such as Nvidia, AWS, Fiddler AI and Hugging Face.
"Domino's approach ... is to make sure that companies and data science teams have access to the latest and greatest innovations," said the vendor's vice president of product marketing, Ramanan Balakrishnan. Partnerships are the best way to harness the innovations and help enterprises manage the different versions of generative AI in the market, he said.
The current market
Domino's release reflects the need among enterprises to take advantage of generative AI and move it from the experimentation or proof of concept (PoC) phase to production simply and in a responsible way, according to industry analysts.
"That's really core to this announcement," said Mike Leone, an analyst with TechTarget's Enterprise Strategy Group (ESG). "It's all around templates that will help organizations build these ML models and utilize generative AI with confidence."
In line with many vendors in the market, Domino recognizes the complexity of generative AI. The benefits of providing templates could help code generation, while data infrastructure will help enterprises jump-start the development and execution of generative AI projects, he added.
Domino is also following other organizations in its strategic partnerships with Nvidia, Fiddler, AWS and Meta, Leone said.
"The partner ecosystem when it comes to generative AI is absolutely critical when organizations are pursuing nuanced use cases in different domains or industries or lines of business," he said.
Leone added that as Domino Data Lab continues to expand its partner ecosystem, it can help organizations looking to apply generative AI outside the common chatbot application. While ESG research has found that chatbot dialogue is the most common use case for generative AI at the moment, other popular applications include text or image generation for marketing, live assist for customer service agents, and suggested scripts for sales.
Shawn RogersAnalyst, BARC
"It's about Domino recognizing that customers will have different requirements depending on the use case, domain or industry, and as such, partnerships will enable their customers to gain the same rapid value across the diverse requirements," Leone said.
Responsible generative AI
Domino is one of many vendors taking a sophisticated approach in its attempt to help enterprises scale their generative AI projects responsibly, said Shawn Rogers, an analyst at BARC.
"That's really going to be the quick art form," he said. "It's one thing to get a PoC off the ground and say, 'Yay, we did a project with AI.' It's another thing to put systems and frameworks in place in your organization that allow you to do it over and over again, in a governed and smart way."
Domino is not the only AI and MLOps vendor taking a responsible AI approach to generative AI. In June, competitor Dataiku released version 12 of its platform, which includes an integration with OpenAI. Dataiku 12 also includes a universal feature that focuses on the explainability of models.
Answering the need for responsible AI, Domino now provides governed access to popular data sources, including new connectivity with more than a dozen sources such as Databricks clusters and SAP.
Domino also introduced into all its workflows new Data Audit Logging, now available for customers. Data Audit enables teams to know where data sources come from and allows for governance across the model lifecycle.
This kind of repeatability, where the right kind of data is framed around the AI models, helps to reduce some of the concerns enterprises have about generative AI, such as hallucinations, Rogers said.
"Companies that reduce risk against making mistakes, risk against falling over or tripping on regulatory things -- those companies are doing a bigger service and delivering greater value to their clients," he said. "Companies that choose the more sophisticated route to helping their clients reduce risk and are able to help them scale -- that's where the winners will be."
Esther Ajao is a TechTarget Editorial news writer covering artificial intelligence software and systems.