Distributional, an innovative AI testing platform founded by Scott Clark, former General Manager of AI Software at Intel, has announced the successful closure of a $19 million Series A funding round, led by Two Sigma Ventures. This funding will enable the company to enhance its services aimed at addressing the growing complexities and risks associated with AI applications.
Inspired by firsthand experiences at Intel and Yelp, where he encountered significant hurdles in AI monitoring and observability, Clark set out to create Distributional. “As the value of AI applications continues to grow, so do the operational risks,” he noted. Distributional’s platform empowers AI product teams to proactively detect, understand, and mitigate risks before they impact production environments.
Clark's journey to Distributional began with Intel’s acquisition of SigOpt, a model experimentation and management platform he co-founded. After becoming VP and GM of Intel’s AI and Supercomputing Software group, Clark realized that AI’s inherent non-deterministic nature made pinpointing bugs akin to finding a needle in a haystack. A staggering 80% of AI projects fail, according to a 2024 Rand Corporation survey, and this challenge is amplified with the rise of generative AI, which Gartner predicts will see a third of its deployments abandoned by 2026.
To combat these issues, Distributional leverages advanced techniques developed while working with enterprise clients at SigOpt. The platform automatically generates statistical tests for AI models and applications tailored to developer specifications, presenting results in an intuitive dashboard. This feature enables teams to collaborate on test repositories, triage failures, and recalibrate as necessary. Furthermore, Distributional can be deployed on-premises or through a managed service, seamlessly integrating with popular alerting and database tools.
Clark emphasizes the platform's ability to provide visibility across organizations, detailing what, when, and how AI applications were tested, and how these parameters have evolved over time. This structured approach to AI testing fosters repeatable processes, utilizing shareable templates, configurations, filters, and tags to enhance efficiency.
Amidst the many AI experimentation solutions on the market, including Kolena, Prolific, Giskard, and Patronus, Clark asserts that Distributional offers a distinct advantage. The company aims to provide a “white glove” experience, managing installation, implementation, and integration for clients, along with troubleshooting support for AI testing.
“Monitoring tools often focus on high-level metrics and specific instances of outliers,” Clark explains, “but they provide a limited understanding of broader application behavior.” Distributional’s goal is to help teams define desired behaviors for their AI applications, ensuring consistency from development through to production and facilitating quick detection of any changes.
With its new funding, Distributional plans to expand its technical team, specifically in UI and AI research engineering. Clark anticipates growing the workforce to 35 by year-end as the startup embarks on its first enterprise deployments.
Having raised a total of $30 million since its inception, including participation from Andreessen Horowitz, Operator Collective, Oregon Venture Fund, Essence VC, and Alumni Ventures, Distributional is poised to seize the immense opportunities in the AI testing landscape. “We have secured significant funding in the course of just a year since we were founded,” Clark stated, “and we are in a position to capitalize on this massive opportunity in the years to come.”
