MCPFlixFree Beta
🧠
intermediateML Engineer

ML Pipeline Starter

Hugging Face + W&B + MLflow + Modal — track experiments, ship models

5 servers~15 min setup

Use Case

Train, track, and deploy ML models without managing infrastructure

The minimal toolkit for ML engineers starting their first production pipeline. Hugging Face for models and datasets, Weights & Biases for experiment tracking, MLflow for model registry, and Modal for serverless GPU compute. Get from notebook to deployed model in days, not months.

Servers in this Stack

Hugging Face

Models & Datasets

Browse models, run inference, and manage datasets on Hugging Face Hub.

Weights & Biases

Experiment Tracking

ML experiment tracking, hyperparameter tuning, and dataset versioning.

MLflow

Model Registry

Open-source platform for managing ML lifecycle including tracking and deployment.

Modal

GPU Compute

Serverless cloud for AI, data, and compute workloads.

GitHub (Official)

GitHub (Official)

Version Control

Repository management, file operations, and GitHub API integration for AI assistants.

One-Click Install

4 of 5 servers will be installed
Paste into ~/.config/Claude/claude_desktop_config.json

How it works

  1. 1.Click "Install Entire Stack" to copy the config
  2. 2.Open your claude_desktop_config.json file
  3. 3.Paste and save
  4. 4.Restart Claude Desktop
  5. 5.Start asking your AI to query data, run dbt models, send Slack messages, etc.