Datarobot MCP integration
Reports deployment accuracy and drift, starts Autopilot runs, queues batch predictions, and keeps datasets linked to the right use case.
836actions available
Three actions you can hand over today
Every action runs live through MCP. Nothing to build, nothing to maintain.
Check project status
Lyro monitors the training and deployment status of DataRobot projects. ML teams avoid deployment surprises by checking health before handoff.
Retrieve model scoring code
Lyro downloads compiled scoring JAR files for production models. Engineers integrate models into applications without waiting for manual uploads.
Get account info
Lyro fetches account details for the authenticated DataRobot user. Teams verify billing and organization settings without navigating dashboards.
How businesses use Datarobot + Lyro
Each card is one request a support team gets, and the Datarobot actions Lyro runs to close it.
Check whether a deployed model is still holding up
Lyro reads accuracy over time, feature drift, and service stats for a deployment, so a question about whether a model has degraded is answered from live monitoring rather than the next scheduled review.
Get Deployment AccuracyList Deployments Feature DriftList Deployments Service StatsStart a modelling run without opening the console
Lyro creates a project from a registered dataset, launches Autopilot against the target you name, and trains a specific blueprint when the team already knows which one it wants.
Create DataRobot ProjectStart DataRobot AutopilotTrain DataRobot ModelKeep datasets registered, refreshed, and filed
Lyro registers a dataset from a URL or connected data source, schedules its refresh, and links it to the use case it belongs to so the workspace stays navigable as projects multiply.
Create Dataset from URLCreate Dataset Refresh JobLink Entity to Use CaseQueue batch scoring and route the alert
Lyro submits a batch prediction job against a deployment, checks the resulting data exports, and sets a notification policy so a failed run reaches the team instead of sitting unnoticed.
Create Batch PredictionsList Prediction Data ExportsCreate Entity Notification Policy
How it works
Get started in 3 steps
Connect once, then just ask. There is no workflow builder to learn and nothing to maintain — Lyro reads the Datarobot actions it has and picks the ones a request needs.
- 01
Connect Datarobot
Authorize the Datarobot account your team already uses — one consent screen, no API keys, no mapping tables. Lyro can only do what you granted that account, and you can disconnect it at any time.
- 02
Tell your agent what you need
Describe the job the way you would hand it to a teammate. Lyro maps it to the Datarobot actions that close it and chains as many as the request needs.
- 03
Watch it work
The agent runs the actions inside the conversation the customer is already in, so nobody copies data between tabs and your team can take over at any point.
Get started free
Everything else about Datarobot
Setup, permissions, and the limits of what Lyro can do inside Datarobot.
Most of the platform. Beyond deployments and monitoring, the toolkit covers projects and models, datasets and wrangling recipes, notebooks and codespaces, GenAI playgrounds and LLM blueprints, registered model packages, and administrative objects such as user groups and access roles. What is reachable in practice is bounded by the permissions on the DataRobot API key you connect.
Every action available in Datarobot
All 836 actions your agent can call on Datarobot, straight from the live MCP connection.
Add users to group
Add one or more users to a DataRobot user group by groupId.
Add user to organization
Add a user to an existing organization.
Analyze dataset definition
Analyze a dataset definition by ID.
Archive model package
Archive a DataRobot model package.
Build java scoring code package
Initiates an asynchronous build of a Java JAR package containing DataRobot Scoring Code for a deployment.
Cancel notebook job
Cancel a running or pending notebook job execution.
Cancel project job
Cancel a pending job for a project.
Check project status
Check the status of a DataRobot project.
Clone application template
Clone an application template into a codespace.
Clone files collection
Create a duplicate files collection in DataRobot.
Copy file or folder
Copy a file or folder within the same DataRobot catalog item.
Create access role
Create a custom Access Role for an organization.
Create and start notebook codespace
Create and start a new notebook codespace in DataRobot.
Create automated document
Request generation of automated compliance documents in DataRobot.
Create batch monitoring job
Create a DataRobot Batch Monitoring job for tracking model performance and data drift on batch predictions.
Create batch monitoring job definition
Create a Batch Monitoring job definition for tracking deployment performance and data drift.
Create batch prediction job definition
Create a Batch Prediction job definition.
Create batch predictions
Create a new batch predictions job in DataRobot.
Create batch predictions from existing
Create a new Batch Prediction job based on an existing job's configuration.
Create batch predictions from job definition
Launch a Batch Prediction job from a job definition.
Create calendar from file upload
Create a DataRobot calendar by uploading a CSV or XLSX file containing date events.
Create calendars from country code
Initialize generation of preloaded calendars from a country code.
Create calendar from dataset
Create a calendar from a dataset in DataRobot.
Create change request
Create a Change Request for a DataRobot deployment to enable governance workflows.
Create code snippets
Generate code snippets for DataRobot models, predictions, or workloads.
Download code snippet
Download code snippets for DataRobot deployments, models, or workloads.
Create comment
Create a comment on a DataRobot entity (deployment, use case, model, catalog, etc.).
Create compliance doc template
Create a new compliance documentation template in DataRobot.
Create credentials
Store a new set of credentials in DataRobot for use with data sources and connections.
Create custom application source
Create a custom application source in DataRobot.
The tools Datarobot sits next to
Same connection, same setup. Pick the next one your team already uses.
Bigml
Inventories models, clusters, and anomaly detectors, reads their evaluations and predictions, and sets up projects and external data connectors.
DeepSeek
Sends prompts to deepseek-chat or deepseek-reasoner with tool calling and thinking mode, and checks balance and model availability first.
Humanloop
Creates a project, reports on its experiments and session history, and deletes the project when the work behind it is finished.
Needle
Searches your document collections and returns ranked passages with sources, uploads new files, retires stale ones, and syncs local connectors.
RunPod
Reports GPU availability and pricing, provisions clusters and serverless endpoints, and stores the registry credentials a private image needs.
Veo
Submits a prompt as a Veo generation job, tracks the operation until it finishes, and downloads the finished clip.
Ready to connect Datarobot?
Authorize the account and your agent has all 836 actions from the first conversation.