Unleash the
Power of AI Oversight

in Databricks with Patronus AI

With a few quick steps, send your MLflow experiment traces from Databricks to Patronus AI’s OTel Collector for effective, detailed insights into your ML workflows.

    Patronus AI

    We are a leading AI evaluation and optimization company. Our research-backed product offerings enable AI engineers to optimize their AI products, access SOTA evaluation models, and automatically detect LLM performance issues across 50+ modes.

    Our platform consists of a suite of features providing end-to-end solution coverage to confidently deploy LLM applications at scale.

    Percival,
    by Patronus AI

    Percival is a SOTA agent evaluator and is capable of detecting 20+ failure modes in agentic traces and suggesting optimizations for agentic systems, spanning reasoning, planning & coordination, and system execution errors.

    Key Benefits

    With this integration with Databricks,
    we leverage the power of Percival to provide:

    1. Real-time Monitoring

    See current, in-progress performance of your MLflow experiments

    2. Evaluation Metrics

    Automatic extraction of performance insights from your traces

    3. Alerting

    Instant and proactive anomaly detection to inform you of any degradation in your workflow

    4. No Extra Setup

    Integrate with provided environment variables with no additional code required

    Get Started Today

    Simple integration documentation to help you get the most out of trace evaluation with Databricks and Patronus AI