Anaconda Looks To Speed AI Development Tasks With New Offering

The new Anaconda AI Catalyst development tool suite includes a catalog of vetted open-source AI models that the company says help bridge the gap between the need for rapid AI application development and compliance with data security and governance requirements.

Anaconda has launched a new suite of development tools that the software development tech company says provides an end-to-end ecosystem for building, deploying and governing AI applications.

With the new AI Catalyst offering, Anaconda—which is well known in the open-source software development space with its Python-based development platform—looks to extend to the AI world its strategy of offering open-source development capabilities combined with data security, governance and compliance.

“Our goal, really in a nutshell, is to do what Anaconda has historically done to make open-source software usable and safe in the Python world—we’re trying to apply that same model to open-source large language models,” said Seth Clark, Anaconda vice president of product, AI, (pictured) in an interview with CRN.

AI Catalyst is an enterprise AI development suite, housed within the broader Anaconda Platform, that provides a curated catalog of secure, vetted, open-source AI models that development teams can use to discover, select, test, compare and run AI models in their own environment, reducing unknown risks and improving governance and cost efficiency, according to the company. AI Catalyst also includes governance tools for managing risk and inference software for running the models.

AI Catalyst, which runs on the Amazon Web Services platform, is designed to address multiple challenges relating to artificial intelligence and software development.

On the more technical side, the “vast majority” of new lines of software code are being written today using a generative AI tool or assistant of some kind, Clark said. While that provides time savings early in the development process, more time is being spent on reviewing and debugging generative AI-created code, adjusting prompts, fixing context windows, and addressing other workflow issues.

Other challenges are on the AI adoption and implementation side, especially in regulated industries like healthcare and financial services with strict rules governing data governance, security and sovereignty issues. That makes things complex for software development teams who are trying to build and deploy AI applications and agents under tight schedules.

“We’re finding that there’s so many rules in place, and these organizations are still figuring out their processes to determine, basically, what amount of risk can they accept,” Clark said. “Where does the data need to live? What models are they going to be willing to work with? What data are those models allowed to touch? There’s a combination of risk management, governance and compliance, legal reviews [and] data sovereignty questions.”

Complying with those requirements and reviews “creates a lot of friction and slows things down,” Clark said. “Our goal would be to help our customers maintain compliance, but with less hurdles, less headaches, and just less time so that engineering teams can get back to doing their jobs.”

“It’s a real wild west in terms of availability, documentation, just getting transparency into what [open-source] models exist, what they’re good for, and which ones are secure,” Clark said. “We’re simplifying that pretty significantly for our customers to have a good, curated set of these models that they can use for those more bespoke use cases where data governance, compliance and risk are pretty significant issues.”

Vetted Open-Source AI Models

The core of AI Catalyst is a set of curated open-source AI models that Anaconda has selected and vetted and come with a “robust” bill of materials, including comprehensive risk profiles for transparency and audit-ready oversight, along with other documentation, according to the company.

The models support local, cloud development or production integrations that run on CPUs or GPUs, “ensuring developers have flexibility when it comes to model deployment within their organization’s secure infrastructure,” according to the Anaconda announcement. The models are optimized and benchmarked for enterprise use cases, which the company says saves developers weeks of manual research and testing, model evaluation, and dependency management, and accelerating prototype-to-production development workflows.

AI Catalyst also provides governance tools that allow organizations to set required risk management levels and data use restrictions. And the platform provides a secure inference server, built using Anaconda’s package distribution, that the company says provides greater control over the inference stack, including supporting verified model execution, thus reducing third-party vulnerabilities and identifying model-specific risks.

“The best way to think about Anaconda’s goal and objective [with AI Catalyst] is we want to provide a really solid foundation that essentially gives engineering teams all of the baseline building blocks they need to create really sophisticated generative AI- and machine learning-powered applications,” Clark said.

Anaconda has seen increasing demand for its development platform from system integrators and consulting firms—mostly those working on data analysis projects for their clients, Clark said. “In all those cases, data sovereignty is one of the biggest challenges they face,” he said.

Partners can deploy the Anaconda platform, including AI Catalyst, within their governed virtual private clouds and safely run AI models where their clients’ data resides without the risk of data exfiltration.

AI Catalyst is currently available on the AWS platform and Anaconda has been demonstrating the product at this week’s AWS re:Invent conference. Clark said it will be available on other platforms by the end of the year and into 2026.

In addition to the release of AI Catalyst, Anaconda also unveiled new capabilities within its core development platform including unified search functionality and expanded model access. Customers also now have the option of running a self-hosted cloud implementation of the Anaconda platform, offered through Amazon Virtual Private Cloud, within their own secure, controlled environments on AWS.

The AI Catalyst launch caps off a busy year for Austin, Texas-based Anaconda. In October the company hired David DeSanto, previously chief product officer at GitLab, as the company’s new CEO. He replaced Anaconda co-founder Peter Wang, the previous CEO, who is now chief AI and innovation officer at the company.

In July, the company raised $150 million in a Series C funding round led by Insight Partners. At the time the company said it had more than $150 million in annual recurring revenue.