DataRobot Inc., which offers tools for developing artificial intelligence and machine learning models that cater to both experts and novices, said today it has snapped up a Toronto-based startup called Agnostiq Inc.

The acquisition will enable DataRobot to facilitate and scale the development of so-called AI agents by offering tools that enhance its compute orchestration and optimization capabilities.

DataRobot is the creator of an AI development platform that’s designed to function as the central hub of a company’s AI initiatives and provides tools for all of the key coding and software deployment tasks involved in creating machine learning applications. What sets DataRobot apart is that it’s targeted at both data scientists and business users. For the latter, it provides a simple point-and-click AI builder that allows them to create AI models without any coding.

It works by evaluating the task that the user wishes to automate with AI, then searches for the most appropriate neural network from its library of prepackaged algorithms. The selected neural network is then trained to carry out that task, with every step fully automated.

For more advanced users, such as expert data scientists, DataRobot offers tools for creating and customizing neural networks. Users can generate a neural network using its no-code AI builder first, and then customize it to their requirements by adding their own code. It’s an approach the company promises is much faster than building everything from scratch.

DataRobot believes its platform also lends itself well to the development of AI agents, which are more advanced forms of AI that can carry out various tasks on behalf of their users with minimal human input or supervision. AI agents have a lot of promise, but DataRobot worries that many companies struggle to manage these applications across fragmented tooling and environments, leading to inefficiencies that slow down innovation. However, it thinks this can be overcome by giving companies the flexibility to deploy agentic AI across multiple infrastructures and compute environments, along with the ability to manage those resources dynamically, based on availability, cost and performance.

That’s where Agnostiq comes in. It’s the creator of an AI infrastructure management and compute orchestration platform called Covalent. This makes it easier to scale AI deployments by utilizing dynamic resource allocation to improve efficiency by deploying them on the most suitable infrastructure.

By integrating DataRobot with Covalent, the company says it can enable heterogeneous compute orchestration so AI agents can be deployed, scaled and managed across any compute environment, including in the cloud or on-premises or even hybrid deployments. Among the capabilities it provides, it helps customers to allocate and burst resources dynamically, so the underlying infrastructure can scale with the best possible price-performance.

In addition, the Covalent platform can help to reduce the complexity associated with the underlying infrastructure, for example by streamlining AI tooling through a single platform without incurring migration costs.

DataRobot Chief Executive Debanjan Saha said while businesses are scrambling to deploy agentic AI workloads, many are held back by the siloed infrastructure and tooling that supports those applications.

“This operational complexity is not sustainable or scalable for business,” he said. “With this acquisition, we are empowering AI teams to intelligently and cost-effectively develop, deliver and govern agentic AI across any compute environment.”

Image: SiliconANGLE/Dreamina