Snorkel AI, Inc.
About Snorkel AI, Inc.
Company overview
Snorkel AI, Inc. is a technology company founded in 2019 by Alexander Ratner, Christopher Ré, Braden Hancock, Henry Ehrenberg, and Paroma Varma. The company is headquartered in Redwood City, California. Spun out of the Stanford AI Lab after several years of research, Snorkel AI focuses on programmatic data development to address the manual data labeling bottleneck in artificial intelligence. The founding team initially led the Snorkel open-source project at Stanford before establishing the commercial entity to meet enterprise demands for scalable AI data solutions. Business/Product Breakdown Snorkel AI, Inc. provides an AI data development platform designed to accelerate the creation, evaluation, and refinement of machine learning models and large language models (LLMs). The company's core product, Snorkel Flow, enables users to replace manual, hand-labeled data processes with programmatic labeling. Through a technique known as weak supervision, subject matter experts write "labeling functions" - such as rules, heuristics, or patterns - that the platform uses to automatically label large, unstructured datasets. This system is designed to reconcile conflicting or noisy signals into high-quality training data. The platform includes tools for guided error analysis, model-agnostic fine-tuning, and a no-code interface for non-technical domain experts to contribute their knowledge. The company also offers specialized services and evaluation frameworks to support production-grade AI. Snorkel Expert Data-as-a-Service provides access to a network of professionals with advanced degrees to generate and validate complex datasets for high-stakes industries such as finance, medicine, and law. Additionally, Snorkel Evaluate allows organizations to assess the performance of frontier models and agentic systems using custom benchmarks and scoring rubrics. The firm's solutions cater to Fortune 500 companies and government agencies, supporting use cases like contract intelligence, news analytics, and the development of specialized AI copilots.
Keywords
Revenue
EBITDA
Valuation
| Line Item | 2024 | 2023 | 2022 | 2021 | 2020 |
|---|---|---|---|---|---|
| Revenues | $380 MM | $250 MM | $90 MM | $120 MM | $65 MM |
| Gross Profit | $190 MM | $125 MM | $45 MM | $60 MM | $32.50 MM |
| Net Income (Loss) | $43 MM | $25 MM | $9 MM | $12 MM | $6.50 MM |
| Line Item | 2024 | 2023 | 2022 | 2021 | 2020 |
|---|---|---|---|---|---|
| Total Assets | $850 MM | $720 MM | $680 MM | $650 MM | $600 MM |
| Total Liabilities | $450 MM | $400 MM | $380 MM | $350 MM | $320 MM |
| Total Equity | $400 MM | $320 MM | $300 MM | $300 MM | $280 MM |
Ratios
Employee Figures
Key Executives
See All401K
M&A2024-03-15Completed
Investor Sponsors NorthBridge PE | Stake Tags 100% |
Deal Amount $120M | Is PE deal Yes |
Target Company Valuation $1.20 MM | Is deal PE backed Yes |
Techniques Acquisition of Assets, Taken Private | Total Acquired 100% |
Investor Sponsors
Deal Amount
$120M
Target Company Valuation
$1.20 MM
Techniques
Acquisition of Assets, Taken Private
Stake Tags
100%
Is PE deal
Yes
Is deal PE backed
Yes
Total Acquired
100%
Secondary Transaction2023-09-01Canceled
Investor Sponsors -- | Stake Tags Majority |
Deal Amount Undisclosed | Is PE deal No |
Target Company Valuation -- | Is deal PE backed Yes |
Techniques Secondary Sale, Management Buy-out (MBO) | Total Acquired 65% |
Investor Sponsors
--
Deal Amount
Undisclosed
Target Company Valuation
NaN
Techniques
Secondary Sale, Management Buy-out (MBO)
Stake Tags
Majority
Is PE deal
No
Is deal PE backed
Yes
Total Acquired
65%
© 2026 PrivCo Holding Inc. All rights reserved.
No AI Training or Automated Data Extraction.
Use of this page or any content or data made available by PrivCo for scraping, crawling, indexing, caching, extracting, copying, reproducing, storing, or for training or developing any artificial intelligence or machine learning system is strictly prohibited without prior written consent. All access to and use of this page and PrivCo's services are subject to the PrivCo Terms of Use: https://www.privco.com/terms-of-use. No license or right is granted by implication or otherwise.