DataRobot Rolls Out New AI App Building Capabilities

'Customers … are not looking for a vector database. They're not looking for millions of parameters and billions of parameters. They’re looking for a solution to their problem,' DataRobot Chief Product Officer Venky Veeraraghavan said.

A new enterprise artificial intelligence suite for composable AI applications and agents, an add-on AI observability feature and enhanced large and unstructured data preparation functionality are some of the latest advancements DataRobot is rolling out to ready partners and customers for the AI era.

The Boston-based AI applications and platform provider revealed the innovations in a series of announcements Tuesday as part of an effort to get AI to end business users, instill more confidence in deploying and running AI apps and unlock more return on organizations’ AI investments, DataRobot Chief Product Officer Venky Veeraraghavan told CRN in an interview.

Making its offerings more modifiable so that solution providers can add in their expertise and build more services on top is an opportunity for the channel, said Veeraraghavan (pictured).

“They can take our general purpose platform, specialize it … and (deliver) an end solution for the customer that has the DataRobot engine inside it, but really it's an application and services that the partner is providing to solve the end problem,” he said. “It really opens up a huge, huge market for partners to build with us and actually then go to market themselves.”

[RELATED: The 20 Hottest AI Software Companies: The 2024 CRN AI 100]

DataRobot AI Enhancements

Customers are less dazzled by the components of an AI product than the results it has, Veeraraghavan said. “They are not looking for a vector database. They're not looking for millions of parameters and billions of parameters. They're looking for a solution to their problem.”

The vendor’s close relationship with SAP could make its platform a welcome addition to those partners looking to improve apps built for Ariba, Concur and other products within the German-based tech giant’s portfolio, he said.

DataRobot has about 300 channel partners worldwide, according to CRN’s 2024 Channel Chiefs.

Enterprise AI Suite

DataRobot’s new enterprise AI suite promises to allow for composable AI apps and agents with pre-built templates for data analysis tools, predictive content creation systems and other AI use cases, according to the vendor.

Users can also adjust security, business and implementation logic of the apps. The suite brings a collaborative AI app library for multiple users in a business to work together on new apps and existing ones.

A GenAI app workshop in the suite allows for rapid prototyping and production deployment with automated monitoring and scaling, according to DataRobot. GenAI action tracing promises to make identifying and resolving root causes simpler.

Users receive out-of-the-box app interface examples from Streamlit, Flask and Slack. They can also use Dash, Shiny, Microsoft Teams and other frameworks for custom interfaces.

Users can tune, refine and view apps in real time, conduct red team activities pre-production and instantly push updates, fixes and improvements without user downtime, according to the vendor.

The suite has a declarative application programming interface (API) framework for faster app development and quick integration with SAP products through DataRobot’s SAP Datasphere connector and SAP AI Core one-click deployment.

More AI Tooling

Along with the enterprise AI suite, DataRobot has more AI tooling for users, including an add-on AI observability feature and one-click compliance documentation.

The full observability wrapper promises to safeguard any app with real-time intervention, moderation and governance in two lines of code, according to DataRobot. The wrapper works for OpenAI, Microsoft Azure, Databricks and other products.

DataRobot also has a new Google Vertex integration so that DataRobot tools and evaluate and monitor products and services powered by Vertex and Google Gemini models.

The one-click AI compliance documentation capabilities covers the European Union’s AI Act, New York’s Law No. 144 and other international, local, and industry regulations, according to DataRobot.

The vendor’s automated compliance tests promise to keep users compliant as new regulations and policies come into effect, and DataRobot offers real-time alerts for compliance.

Unstructured Data Updates

DataRobot enhanced its large and unstructured data preparation and handling functionality to better support GenAI and predictive AI development, according to the vendor.

The new and enhanced functionality includes the ability to automate data quality assessments, remediation and healing as well as relationship detection. The detection automation capability allows users to identify and join similar features across datasets and find new features.

Users can leverage built-in optical character recognition (OCR) to make unstructured documents ready for AI and build vector databases for faster data retrieval. They can also improve response accuracy with indexing embeddings with metadata, semantically chunking text, enabling multiple retrievals for complex queries, rewriting queries based on chat history, and other techniques.

The DataRobot updates will also allow users to process data in iterations, avoid overfitting on large datasets, connect to Nvidia’s Rapids AI libraries and other tools and bring in Nvidia Nim microservices, according to DataRobot.