WhyLabs

What do you do?

WhyLabs is the essential AI Observability Platform for model health and data health. Enterprise data teams use the platform to monitor data pipelines and AI applications, to surface and resolve data quality issues, data bias and concept drift. These capabilities help AI builders reduce model failures, avoid downtime, and ensure customers are getting the best user experience. 

With out-of-the-box anomaly detection and purpose-built visualizations, WhyLabs eliminates the need for manual troubleshooting and reduces operational costs. The platform works on any data, structured or unstructured, at any scale, on any cloud platform. The company was created at a fundamental AI research institute – the Allen Institute for Artificial Intelligence (AI2) – by Amazon Machine Learning alums.

How much does it cost?

Free tier
per model per month

Video/Tutorials

The WhyLabs Platform is available for anybody to experience through the sandbox: try.whylabsapp.com

What’s a sample use case? Where can I learn from?

Feature List

  • Monitor for data drift, data quality and availability across structured data, images, text, and audio
  • Monitor model performance with turn-key or customizable accuracy metrics 
  • Configure monitoring in minutes with zero-configuration setup, learned thresholds and complex baselines
  • Receive alerts in your current workflow, i.e. Slack, PagerDuty, ServiceNow, etc.
  • Integrate quickly through lightweight, privacy-preserving open source data logging library – whylogs 
  • Switch on monitoring for batch and streaming pipelines with a few lines of code across popular data tools: Spark, Kafka, MLFlow, Sagemaker, etc
  • Analyze historical feature distributions and model performance with unlimited data retention 
  • Debug data and model issues easily with purpose-built visualizations and rich user interface