Making practical ML-driven systems trustworthy with foundational research.
The heart of AI Quality efforts
We should be able to trust the models we build. We work on AI Quality analytics: enabling the deep evaluation and introspection of models that is necessary to build high-quality, trustworthy ML systems. This involves properly explaining, debugging, and monitoring models and their data.
Approach and philosophy
We work on real-world problems and translate fundamental research discoveries into product capabilities.