author
Co-founder
,
Snorkel AI
Henry Ehrenberg is a co-founder of Snorkel AI, focused on technical strategy and engineering. He has been a core Snorkel team member since the project’s origins in the Stanford AI Lab, building the open-source research library and conducting research on programmatic data labeling and augmentation.
Before Snorkel AI, Henry was the tech lead for Facebook Applied AI’s representation learning team. Henry earned his master’s degree in computational and mathematical engineering from Stanford University, and his bachelor’s degree in applied mathematics from Yale University.
The latest from Henry
Blog
Snorkel AI Teams with Google Cloud and Vertex AI to speed AI deployment
Snorkel AI, Google Cloud and Vertex AI partner to help organizations transform data into AI-powered systems faster than ever.
Blog
Deepening Snorkel AI’s partnership with Microsoft Azure AI
Snorkel AI is excited to build on our partnership with Microsoft Azure to help enterprises and government agencies solve their most impactful problems and unlock value from their data using AI. Learn how Azure customers can easily deploy Snorkel Flow on their Azure cloud infrastructure to accelerate AI application development with data-centric workflows and programmatic labeling.
Blog
Q4 LTS Release of Snorkel Flow
We’re excited to announce the Q4 2021 LTS release of Snorkel Flow, our data-centric AI development platform powered by programmatic labeling. This latest release introduces a number of new product capabilities and enhancements, from a streamlined programmatic data development interface, to enhanced auto-suggest for labeling functions, to new machine learning capabilities like AutoML, to significant performance enhancements for PDF data…
Snorkel: Fast Training Set Generation for Information Extraction
State-of-the art machine learning methods such as deep learning rely on large sets of hand-labeled training data. Collecting training data is prohibitively slow and expensive, especially when technical domain expertise is required; even the largest technology companies struggle with this challenge. We address this critical bottleneck with Snorkel, a new system for quickly creating, managing, and modeling training sets. Snorkel enables users to generate large volumes of training data by writing labeling functions, which are simple functions that express heuristics and other weak supervision strategies. These user-authored labeling functions may have low accuracies and may overlap and conflict, but Snorkel...
Research Paper
Snorkel: Fast Training Set Generation for Information Extraction
State-of-the art machine learning methods such as deep learning rely on large sets of hand-labeled training data. Collecting training data is prohibitively slow and expensive, especially when technical domain expertise is required; even the largest technology companies struggle with this challenge. We address this critical bottleneck with Snorkel, a new system for quickly creating, managing, and modeling training sets. Snorkel…
Learning to Compose Domain-Specific Transformations for Data Augmentation
Data augmentation is a ubiquitous technique for increasing the size of labeled training sets by leveraging task-specific data transformations that preserve class labels. While it is often easy for domain experts to specify individual transformations, constructing and tuning the more sophisticated compositions typically needed to achieve state-of-the-art results is a time-consuming manual task in practice. We propose a method for automating this process by learning a generative sequence model over user-specified transformation functions using a generative adversarial approach. Our method can make use of arbitrary, non-deterministic transformation functions, is robust to misspecified user input, and is trained on unlabeled data....
Research Paper
Learning to Compose Domain-Specific Transformations for Data Augmentation
Data augmentation is a ubiquitous technique for increasing the size of labeled training sets by leveraging task-specific data transformations that preserve class labels. While it is often easy for domain experts to specify individual transformations, constructing and tuning the more sophisticated compositions typically needed to achieve state-of-the-art results is a time-consuming manual task in practice. We propose a method for…
Snorkel: Rapid Training Data Creation With Weak Supervision
Labeling training data is increasingly the largest bottleneck in deploying machine learning systems. We present Snorkel, a first-of-its-kind system that enables users to train state-of-the-art models without hand labeling any training data. Instead, users write labeling functions that express arbitrary heuristics, which can have unknown accuracies and correlations. Snorkel denoises their outputs without access to ground truth by incorporating the first end-to-end implementation of our recently proposed machine learning paradigm, data programming. We present a flexible interface layer for writing labeling functions based on our experience over the past year collaborating with companies, agencies, and research labs. In a user...
Research Paper
Snorkel: Rapid Training Data Creation With Weak Supervision
Labeling training data is increasingly the largest bottleneck in deploying machine learning systems. We present Snorkel, a first-of-its-kind system that enables users to train state-of-the-art models without hand labeling any training data. Instead, users write labeling functions that express arbitrary heuristics, which can have unknown accuracies and correlations. Snorkel denoises their outputs without access to ground truth by incorporating the…
Data Programming With DDLite: Putting Humans in a Different Part of the Loop
Populating large-scale structured databases from unstructured sources is a critical and challenging task in data analytics. As automated feature engineering methods grow increasingly prevalent, constructing sufficiently large labeled training sets has become the primary hurdle in building machine learning information extraction systems. In light of this, we have taken a new approach called data programming [7]. Rather than hand-labeling data, in the data programming paradigm, users generate large amounts of noisy training labels by programmatically encoding domain heuristics as simple rules. Using this approach over more traditional distant supervision methods and fully supervised approaches using labeled data, we have been...
Research Paper
Data Programming With DDLite: Putting Humans in a Different Part of the Loop
Populating large-scale structured databases from unstructured sources is a critical and challenging task in data analytics. As automated feature engineering methods grow increasingly prevalent, constructing sufficiently large labeled training sets has become the primary hurdle in building machine learning information extraction systems. In light of this, we have taken a new approach called data programming [7]. Rather than hand-labeling data,…
