Distributional Aims To Develop Software To Reduce AI Risk


Companies are increasingly interested in harnessing the potential of AI to enhance productivity, but concerns about the associated risks continue to loom large. According to a recent Workday survey, enterprises identify the timeliness and reliability of underlying data, potential bias, and security and privacy as the primary barriers to AI implementation.

Key Takeaway

Distributional, led by Scott Clark, is developing a modern enterprise platform for AI testing and evaluation to proactively identify and address AI risk before it impacts customers in production.

Scott Clark’s Vision for Distributional

Sensing a business opportunity in addressing these concerns, Scott Clark, the co-founder of the AI training and experimentation platform SigOpt (acquired by Intel in 2020), has embarked on a mission to develop software aimed at making AI safe, reliable, and secure. Clark’s new venture, Distributional, is focused on creating a platform to standardize and scale tests for various AI use cases.

The Need for AI Testing and Evaluation

Clark’s motivation to launch Distributional stemmed from his experiences at Intel, where he encountered challenges related to AI after the SigOpt acquisition. He observed the difficulty in ensuring high-quality AI testing on a regular basis, leading him to recognize the need for comprehensive AI testing and evaluation.

Distributional’s Approach

The core product of Distributional is designed to detect and diagnose potential harm from large language models and other AI models, offering organizations a comprehensive view of AI risk in a pre-production environment. The platform includes an extensible testing framework, a configurable testing dashboard, and an intelligent test suite to continuously analyze stability and robustness.

Competing in the Market

Despite being pre-revenue and pre-launch, Distributional aims to differentiate itself through its enterprise-focused software, tailored to meet the data privacy, scalability, and complexity requirements of large enterprises across regulated and unregulated industries. The company is currently in the process of co-designing the product with enterprise partners.

Future Outlook

Distributional anticipates generating revenue once its platform launches in general availability, with plans to secure paid customers among its design partners. The company recently closed an $11 million seed round led by Andreessen Horowitz’s Martin Casado, signaling confidence from investors in Distributional’s vision.

Leave a Reply

Your email address will not be published. Required fields are marked *