Artificial intelligence is proving to be transformational for companies across industries and use cases, but the risk associated with machine learning model failure is inhibiting its potential. To avoid risk and send models into production safely, data scientists and ML engineers need to test and continuously validate models and data, and effectively alert and communicate model failures. As data science organizations are shipping models to production, prevention of risk associated with model failures and ensuring ML integrity becomes a top priority.

This is why we are so excited to partner with Databricks, the leading lakehouse platform. Robust Intelligence has integrated our end-to-end ML integrity platform with Databricks to accelerate the world’s best data science organizations by proactively mitigating model failure. Databricks streamlines the entire data science workflow — from data prep to modeling to sharing insights — with a collaborative and unified data science environment built on an open lakehouse foundation. Robust Intelligence will build on the lakehouse platform to enable customers to accelerate AI adoption.

Models fail due to a variety of reasons. Corrupted data, model drift, biased decisions, liabilities, software supply chain vulnerabilities, and adversarial input can all cause model failures. These failures can have dire consequences when the models are used to make business-critical decisions, from credit worthiness and healthcare coverage to hiring practices and aid allocation. Leading companies are adopting strategies and engineering paradigms to ensure integrity into their ML systems. ML integrity is the assurance of data quality, model performance, fairness, security, and transparency. It is only achieved when such considerations are applied to every stage of the ML lifecycle - from model development to production.

By running the Robust Intelligence platform natively within Databricks, customers are able to seamlessly scale their MLOps pipeline while ensuring trust in the system. Robust Intelligence performs automated stress testing to models and data in development to ensure that all vulnerabilities have been eliminated. Once in production, the AI Firewall protects models from bad data in real time while continuous testing ensures that models remain valid over time. The platform performs these computations in a Databricks cluster, which quickly scales to meet the large data sets typical of enterprise machine learning projects.

“We are seeing rapid adoption of a lakehouse by companies that are forward-thinking about machine learning and AI,” said Roger Murff, VP of Technology Partners at Databricks. “Machine learning integrity through continuous testing is one of the keys to their success. We're excited to partner with Robust Intelligence on the Databricks Lakehouse Platform to enable customers to fully realize the value of machine learning and AI."

Native integrations with the Databricks Lakehouse Platform enable Robust Intelligence users to connect directly to their lakehouse, and further integrations with Databricks’ hosted MLflow solution will enable users to track experiments and ingest models into the Robust Intelligence platform directly from the Model Registry just by providing its uniform resource identifier (URI).

Using Robust Intelligence and Databricks together puts an end to 2 AM firefighting calls, endless hours customizing your own internal tooling, and silent errors bringing risk to your business and customers. We instill trust in ML models, enabling organizations to develop intelligent applications with confidence and at greater velocity.

Get Started

Just sign-up for a free trial of Robust Intelligence. Our integration is designed to support any number of models, from one to hundreds, across tabular, NLP, and CV modalities. Customers can rest assured that their data and models are kept secure, as only test results are sent to the Robust Intelligence control plane and the platform is certified SOC 2 compliant.