Getting Started

A guide to using Monitor ML APIs.

Projects

Store your projects in MonitorML to group models and track version changes in models

Add Project

The request body is in JSON format.

  • Request Type: Post
  • Request Body
    • name, Type: String

app.monitorml.com/api/v1/projects

import requests
import json
requests.post("https://app.monitorml.com/api/v1/projects",
  data=json.dumps({name: "Image Recognition"}),
  headers={
    "x-api-key": "YOUR_API_KEY",
    "Content-Type": "application/json"
  }
)
                

Get All Projects

  • Request Type: Get

app.monitorml.com/api/v1/projects

import requests
requests.get("https://app.monitorml.com/api/v1/projects",
  headers={
    "x-api-key": "YOUR_API_KEY",
  }
)
                

Models

Store your models in MonitorML to quickly track model deployment status, model metadata, and latest model metrics and events.

Add Model

The request body is in JSON format.

  • Request Type: Post
  • Request Body
    • name, Type: String
    • status, Type: string
    • metadata_json, Type: JSON
    • project_uuid, Type: String

app.monitorml.com/api/v1/mlmodels

import requests
import json
requests.post("https://app.monitorml.com/api/v1/mlmodels",
  headers={
    "x-api-key": "YOUR_API_KEY",
    "Content-Type": "application/json"
  },
  data=json.dumps({
    "status": "Training",
    "metadata_json": {},
    "name": "facial-recognition-1",
    "project_uuid": "PROJECT_UUID"
  })
)
                

Get All Models

  • Request Type: Get

app.monitorml.com/api/v1/mlmodels

import requests
requests.get("https://app.monitorml.com/api/v1/mlmodels",
  headers={
    "x-api-key": "YOUR_API_KEY",
  }
)
                

Get a Model

  • Request Type: Get

app.monitorml.com/api/v1/mlmodels/your_model_uuid

import requests
requests.get("https://app.monitorml.com/api/v1/mlmodels/MODEL_UUID",
  headers={
    "x-api-key": "YOUR_API_KEY",
  }
)
                

Update a Model

The request body is in JSON format.

  • Request Type: Put
  • Request Body
    • name, Type: String
    • status, Type: string
    • metadata_json, Type: JSON
    • project_uuid, Type: String
    • archived, Type: Boolean

app.monitorml.com/api/v1/mlmodels/your_model_uuid

import requests
import json
requests.post("https://app.monitorml.com/api/v1/mlmodels/MODEL_UUID",
  headers={
    "x-api-key": "YOUR_API_KEY",
    "Content-Type": "application/json"
  },
  data=json.dumps({
    "status": "Training",
    "metadata_json": {},
    "archived": True
  })
)
                

Metrics

Store your metrics in MonitorML to track model performance over time.

Add Metrics

The request body is in JSON format.

  • Request Type: Post
  • Request Body
    • mlmodel_uuid, Type: String
    • status, Type: String
    • metrics, Type: JSON

app.monitorml.com/api/v1/metrics

import requests
import json
requests.post("https://app.monitorml.com/api/v1/metrics",
  headers={
    "x-api-key": "YOUR_API_KEY",
    "Content-Type": "application/json"
  },
  data=json.dumps({
    "mlmodel_uuid": "Model UUID (found on dashboard)",
    "status": "Training",
    "metrics": {
      "accuracy": 0.90,
      "prediction_time": 1.1
    },
    "step": {
      "gradient_descent_iterations": 50
    }
  })
)
                

Events

Store your events in MonitorML to analyze events over time.

Today you can add your events below and each model will have the average value for the numeric metrics calculated and aggregated per day. View it by clicking "Events" underneath the model's graph on the dashboard

Add Events

The request body is in JSON format.

  • Request Type: Post
  • Request Body
    • mlmodel_uuid, Type: String
    • event_data, Type: JSON

app.monitorml.com/api/v1/events

import requests
import json
requests.post("https://app.monitorml.com/api/v1/events",
  headers={
    "x-api-key": "YOUR_API_KEY",
    "Content-Type": "application/json"
  },
  data=json.dumps({
    "mlmodel_uuid": "Model UUID (found on dashboard)",
    "event_data": {
      "conversion": 1
    }
  })
)
                

How to get support

MonitorML provides support for bugfixes and guidance on using the platform.

To access support, send an email to founders(at)monitorml.com

Giving Feedback

We strive to improve our products and we rely on feedback from our customers.

Please feel free to share any feedback about MonitorML to founders(at)monitorml.com.