MLflow compatible. Zero friction.

Stop experimenting.
Start shipping.

Drop-in enhancement for MLflow that adds deployment capabilities and a modern UI. Deploy ML models in minutes, not weeks.

Get StartedSee it in action
Ready to use
Open source
pip install mltrack
mltrack-demo

Start building in seconds

MLTrack works with your existing MLflow setup. No migrations, no breaking changes.

Quick Start

Get started with MLTrack in seconds

$pip install ml-track

ML deployment is unnecessarily hard

You've built a great model. Don't let deployment hold you back.

Weeks to deploy

Traditional ML deployment involves containers, APIs, infrastructure setup, and endless configuration.

MLflow's dated UI

MLflow's interface feels like 2015. Finding experiments, comparing models, and tracking costs is painful.

Complex workflows

Jumping between notebooks, MLflow UI, cloud consoles, and deployment scripts breaks your flow.

No quick testing

Want to test a model? Set up an API, write client code, handle authentication... hours later.

MLTrack makes it delightfully simple

Keep using MLflow. Add MLTrack. Ship models in minutes.

Before MLTrack:
mlflow ui
→
With MLTrack:
ml ui

Beautiful, modern interface that actually helps you work

Before MLTrack:
Complex Docker setup + API code + Cloud config
→
With MLTrack:
ml ship model --modal

One command to deploy anywhere

Before MLTrack:
MLproject files + conda.yaml + setup.py
→
With MLTrack:
ml train script.py

Just run your script, we handle the rest

Before MLTrack:
Write test client + Handle auth + Parse responses
→
With MLTrack:
ml try model --modal

Test deployed models instantly

100% compatible with your existing MLflow setup

No migrations
No breaking changes
Just enhancement

Everything you need to ship ML models

Built by ML engineers who got tired of the deployment dance.

One-command deployment

ml ship model --modal deploys to production. No Dockerfile, no YAML, no tears.

Beautiful modern UI

Finally, an ML dashboard that doesn't look like it's from 2015. Dark mode included.

Deploy anywhere

Modal, AWS Lambda, or Docker. More platforms coming soon. Your models, your choice.

CLI that sparks joy

Intuitive commands that make sense. ml train, ml save, ml ship. That's it.

Cost tracking built-in

See exactly how much each model costs to train and deploy. No surprises.

Team-friendly

Share models, track experiments, and collaborate without the complexity.

MLflow compatible

Works with your existing MLflow setup. No migrations, no data loss, just enhancement.

Lightning fast

Modern tech stack means instant loads, real-time updates, and zero lag.

Ready to ship your models?

Join the growing community of ML engineers who ship models in minutes, not weeks.

Get Started on GitHub
pip install mltrack