Robust Intelligence Raises $30M Series B to Stress Test AI Models

Robust Intelligence Secures $30 Million in Series B Funding

Robust Intelligence, an AI-focused startup dedicated to enhancing the reliability of AI models and preventing failures, has announced the successful completion of a $30 million Series B funding round. This investment was spearheaded by Tiger Global.

Existing investor Sequoia, who previously led the company’s Series A funding, participated in this oversubscribed round. Additional participation came from Harpoon Venture Capital and Engineering Capital.

Company Origins and Leadership

The company was established by Yaron Singer, a tenured professor specializing in Computer Science and Applied Mathematics at Harvard University. He collaborated with his former student, Kojin Oshiba, to co-found the venture.

Robust Intelligence CEO Yaron Singer. Image Credits: Robust Intelligence

The Evolution of AI and the Need for Rigor

Singer commented on the transformation of AI, stating it has evolved from an academic pursuit to a practical reality. He highlighted how advancements in internet infrastructure, data availability, and processing power have unlocked AI’s potential within a relatively short timeframe.

He further explained the necessity of adopting a level of rigor comparable to that used in traditional software development, a practice refined over six decades. The goal is to address the unique challenges presented by AI and ensure its dependability.

Addressing the Inherent Uncertainty in AI

As Singer pointed out, the statistical foundation of AI can lead to unpredictable outcomes. Consequently, Robust Intelligence’s core mission is to mitigate these errors and enhance the overall robustness of AI systems.

Introducing the Robust Intelligence Model Engine (RIME)

To achieve this, the company provides its users with the Robust Intelligence Model Engine (RIME). This system functions as an AI firewall, safeguarding AI models from errors through continuous and comprehensive stress testing.

“With a single click, users can initiate stress testing on their AI models and associated data,” Singer stated. This process encompasses both pre-production evaluation and ongoing monitoring while the model is operational.

The system is designed to automatically identify potential failure points within a model and detect issues such as data drift.

Image Credits: Robust Intelligence

An AI Firewall Powered by AI

Notably, the AI firewall itself is an AI model designed to predict whether a specific data point will result in an incorrect prediction. Singer emphasized that this represents a significant challenge within the fields of AI and machine learning.

Investor Perspective

John Curtius, a partner at Tiger Global, shared his initial exposure to Robust Intelligence’s capabilities during the company’s early stages. He noted the substantial growth of both the company and its product over the past year.

Curtius expressed confidence that Robust Intelligence is fundamentally changing the landscape of AI reliability and that Tiger Global is well-positioned to provide crucial resources to support its continued success.

Future Plans

The newly acquired funding will be allocated primarily to expanding the company’s product development and engineering teams. A portion will also be used to scale its sales operations.