How Snorkel’s Obsession with Data Fuels Its Success in Enterprise AI

Snorkel AI’s journey began when Alex Ratner, co-founder and CEO, shifted his focus at Stanford’s AI Lab from improving machine learning models to solving data challenges after seeing that lack of quality training data was a major bottleneck for real-world AI.

Initially considered a low-priority task, data labeling became the core challenge Snorkel AI aimed to address. With co-founders and early projects focused on developing tools for programmatically labeling training data, Snorkel eventually gained traction, attracting major industry players who needed efficient data processing solutions, especially for complex documents.

The company’s early success came from partnering with large enterprises, such as banks and healthcare institutions, that needed specialized AI models to process proprietary data accurately. These projects, like a major U.S. bank’s LIBOR transition, demonstrated the importance of high-quality, use-case-specific data in AI development. Snorkel’s approach, driven by customer feedback and real-world applications, laid the foundation for Snorkel Flow, a platform now used across industries to reduce the time needed for preparing data from months to weeks. The arrival of ChatGPT initially disrupted AI’s strategic focus, but Snorkel weathered the hype cycle, ultimately reinforcing the demand for AI solutions tailored to enterprise-specific data.

Under Ratner’s leadership, Snorkel has maintained a customer-centric and data-focused ethos, navigating rapid growth while keeping true to its founding insights. Ratner has adapted his role from hands-on development to high-level strategic focus, empowering his team while diving into specific areas where he can contribute most. This dual approach—deep involvement in pivotal areas and empowering team autonomy—has helped Snorkel grow to a 190-person company with an unwavering focus on delivering AI models that align with customer data needs and business goals.