Interested in building the future of machine learning? We’re hiring!See our open roles 👾

ML, without the FML

Instant serverless machine learning environments with one command. Code locally, run in the cloud.


Lightning fast setup

Get up and running with ML projects with near-zero setup

  • Setup with one command

    All you need to do is run unweave init. Unweave does the rest to instantly connect your local code to remote infra.

  • Clone code and data seamlessly

    Unweave makes cloning and sharing datasets as easy as cloning a Git repo. Two clicks and you’re done.

  • Run with remote GPUs

    Unweave gives you and your team instant access to serverless GPU compute in the cloud. No setup required.

  • Scale your cloud infra instantly

    Your code runs in the cloud by default which means that your dev environments are always in sync across your team.

  • Sync code and data on Git

    Your environment and code are always in sync. Everything in Unweave is versioned with Git.

  • Effortlessly manage access

    Nobody likes configuring IAM policies. Unweave manages permissions at a GitHub repository level.

Integrate, without losing your mind

Unweave makes it simpler and easier than ever to get your machine learning projects moving at light speed


Instant cloud infrastructure

No more manually configuring cloud GPU instances or storage: Unweave integrates with cloud providers to automatically provision the compute your projects need, without delays or complex workflows.


Effortlessly add collaborators and manage project permissions

Unweave uses your existing GitHub repositories to manage access to data and compute, so you don’t need any additional IAM or permissions


Get started with an Unweave template

Whether you’re starting your first project, or streamlining your team’s ML workflow, Unweave makes it swift and seamless

  • COCO

    Common Object in Context. COCO is a large-scale object detection, segmentation, and captioning dataset.


    The MNIST database of handwritten digits. Digits have been size-normalized and centered in a fixed-size image.

  • GLUE

    GLUE benchmark is a collection of resources for analyzing natural language understanding systems.

  • IMDb Movie Reviews

    The IMDb Movie Reviews dataset is a binary sentiment analysis dataset of reviews as positive or negative.

  • Kinetics Human Action Video

    A large-scale, high-quality dataset for human action recognition in videos. Each video clip lasts around 10 seconds.

  • Stanford Dogs Dataset

    The Stanford Dogs dataset contains images of 120 breeds of dogs from around the world.

Interested in building the future of machine learning? Join us! 👾