Image

Anomaly Detection

Rapidly build high-accuracy, adaptive anomaly detection models over documents, logs, network traffic, and more, without hand-labeling training data using Snorkel Flow.

Request demo

Image
Image
8 MONTHS

of hand labeling replaced in a few hours using Snorkel

Overview

Detect Needles in the Data Stack

Rapidly build, adapt, and deploy agile anomaly detection applications.

Image

Faster, Lower-cost Development

Use programmatic labeling to develop high-quality AI applications in hours instead of spending weeks or months on expensive hand-labeling.

Image

Rapidly Adaptable

Monitor for changes in the data, and rapidly adapt using built-in error analysis tools. Zoom in on errors to fine-tune training data & models with guided iteration.

Image

High-accuracy Models

Leverage large amounts of labeled and unlabeled data, and state-of-the-art anomaly detection model architectures to build high-accuracy anomaly detectors.

Image

Flexible Integrations

Easily integrate labeling, training, and analysis pipelines defined over diverse input types–text, PDF, HTML, and more–with downstream applications using APIs or a Python SDK.

Use Cases

Explore Enterprise Solutions For Anomaly Detection

Case Study

Image
Stanford Medicine used Snorkel to detect medical anomalies using cross-modal data from radiographs, tomographs & monitoring devices, replacing months of hand labeling with few hours of programmatic data creation.
Problem

Labeling training data for triaging models took person-months to person-years of radiologist time.

Solution

Developed a cross-modal, anomaly detection application using Snorkel, matching or exceeding the performance of painstakingly gathered manual labels in hours.

Result

Currently being tested for deployment in Stanford Health & Department of Vetaran Affairs (VA) hospital systems.

8 MONTHS

of hand-labeling replaced by a few hours

94%

ROC AUC Performance

50K+

images labeled in minutes

How Snorkel Flow Works

Build Agile AI Apps for Anomaly Detection

Build agile anomaly detection applications that identify even rare occurrences in mountains of data. Eliminate onerous hand-annotation processes by programmatically labeling data with powerful labeling functions. Adjust data sampling to zero in on a rare class and augment data using class-aware tooling. Train state-of-the-art models and analyze performance using flexible tools for class-imbalanced problems. Deploy and integrate anomaly detection applications with other data tasks.

An End-to-end ML Platform

Designed for Collaboration

Image

Data Scientist Friendly

  • Integrated Jupyter notebooks
  • Instant analysis tools
  • Ready-to-use models
Image

Domain Expert Friendly

  • Intuitive, no-code UI
  • Rich dashboards and visualizations
  • Full-featured, push-button error analysis
Image

Developer Friendly

  • Platform access via Python SDK
  • Online or batch API deployment
  • Containerized software for cloud or on-premises deployments

Research

Based on Years of Novel Research
Learn more about groundbreaking techniques for machine learning and weak supervision developed by the Snorkel AI team at Stanford AI Lab and beyond.