Bring Ml Lifecycle
to Google ADK

Connect your MLflow tracking server to any AI agent and take full control of your machine learning experiments, training telemetry, and model registry through natural conversation.

What you can do

• Run Orchestration — Search and retrieve detailed Model Training Runs across specific experiments to track accuracy metrics, loss curves, and scalar parameters directly from your agent
• Experiment Audit — List all registered MLflow experiments and retrieve detailed metadata configurations to understand how your project's research branches are structured
• Metric Inspection — Extract explicit telemetry capturing the exact state vectors and performance metrics logged during atomic training sessions for rapid diagnostic analysis
• Model Registry Management — Search the Global Model Registry to identify models explicitly promoted to production or staging pipelines and track version deployments securely
• Artifact Visibility — List physical storage boundaries referencing stored model blobs, image graphs, or metadata saved natively inside MLflow training runs
• Telemetry Mapping — Aggregate tracking logs from multiple experiments to identify trends and compare model performance across different historical training sessions

How it works

1. Subscribe to this server
2. Enter your MLflow Tracking URI and Tracking Token
3. Start managing your ML experiments from Claude, Cursor, or any MCP-compatible client

Who is this for?

• Data Scientists — monitor training progress and verify model metrics through natural conversation without manual dashboard navigation
• ML Engineers — audit the model registry and verify artifact storage locations directly from your workspace terminal
• AI Operations Teams — track production model versions and ensure consistent deployment of high-performing ML models efficiently