Snorkel AI, Inc.
Snorkel AI is a frontier AI data lab that helps enterprises and frontier AI teams build specialized, production-grade AI using expert data development. It creates datasets, environments, evaluation frameworks, benchmarks, and custom solutions for real-world AI systems.
At a Glance
- Frontier AI labs and model developers
- Fortune 500 and Global 2000 enterprises
- Financial services and banking
- Telecommunications
- +5 more
AI Tools by Snorkel AI, Inc.
(1)
Snorkel AI
AI Training Data and Evaluation
Discussions
No discussions yet
Be the first to start a discussion about Snorkel AI, Inc.
Latest News
Expanded Open Benchmarks Grants commitment 10x to $30M for open frontier-AI evaluation research.
Raised $350M Series E at a $3.5B valuation, co-led by Insight Partners and S32.
Highlighted the first wave of Open Benchmarks Grants projects.
Snorkel AI, Princeton and UW-Madison researchers released Senior SWE-bench.
Products & Services
Enterprise AI data-development platform for programmatic labeling, flexible model training, iterative error analysis and improvement, and deployment of custom AI applications; available as hosted or on-premises software.
Visual/no-code builder within Snorkel Flow with pre-built templates and reusable workflows for use cases such as contract intelligence, news analytics, customer-interaction routing, classification, NER and information extraction.
Evaluation platform for building specialized, scalable benchmark datasets and expert-aligned evaluations for enterprise AI systems and agents.
White-glove, expert-produced data service for evaluation and post-training datasets, including demonstrations, reasoning traces, preference labels, rubrics and programmatic checks.
Market Position
Snorkel positions itself as a research-led, data-centric AI data lab rather than a generic labeling vendor: it combines programmatic data development, expert human input, model evaluation, benchmarks and agentic environments with an enterprise platform and delivery services. Its competitive set includes Labelbox, Scale AI, Databricks and other AI data/evaluation platforms; its differentiation is specialized expert data, workflow integration and research-backed evaluation for difficult, domain-specific AI tasks.
Leadership
Founders
Alexander (Alex) Ratner
Co-founder and CEO; completed a Stanford computer-science PhD advised by Christopher Ré, started and led the Snorkel open-source project, and is an affiliate assistant professor of computer science at the University of Washington.
Braden Hancock
Co-founder; previously Director of AI at Meta.
Christopher Ré
Co-founder and Stanford AI Lab professor/researcher; advised Alex Ratner's PhD and helped develop the original Snorkel research program.
Henry Ehrenberg
Co-founder; a member of the Stanford-originated research team that spun Snorkel AI out of the Stanford AI Lab.
Paroma Varma
Co-founder and Head of Research; leads research at Snorkel AI and is part of the Stanford-originated founding team.
Manas Joglekar
Co-founder and member of the Stanford AI Lab-originated Snorkel founding team.
Vincent Sunn Chen
Co-founder and Vice President; part of the founding Snorkel AI team and currently serves as a company vice president.
Executive Team
Alexander (Alex) Ratner
Co-founder and CEO
Stanford computer-science PhD; started and led the Snorkel open-source project; affiliate assistant professor at the University of Washington.
Jimmy Hillis
Chief Technology Officer
Joined Snorkel AI as CTO in 2025 after prior technology leadership work; publicly described returning to San Francisco for the role.
Board of Directors
Founding Story
Snorkel began with a contrarian Stanford AI Lab thesis that training data, rather than only models or compute, would determine whether machine learning worked. The team saw that manual labeling did not scale for data-hungry, specialized and privacy-sensitive domains, developed programmatic data-labeling methods through the open-source Snorkel project, and spun out in 2019 to make that approach usable through Snorkel Flow.
Business Model
Revenue Model
Enterprise software subscriptions/licenses for Snorkel Flow, deployed hosted or on-premises under customer order forms, plus enterprise services and expert Data-as-a-Service for custom evaluation and post-training data. Sales are primarily demo- and contract-led rather than self-serve.
Target Markets
- Frontier AI labs and model developers
- Fortune 500 and Global 2000 enterprises
- Financial services and banking
- Telecommunications
- Insurance
- Biotechnology and healthcare
- Agentic AI and multi-step reasoning
- Coding and software engineering
- Computer-use agents
- Scientific research workflows
- Legal research and case-law analysis
- Insurance underwriting
- Wayfair
- Chubb
- BNY