About Ivy

At Ivy, we’re on a mission to unify all Machine Learning (ML) frameworks, making ML code cleaner, more flexible, and fully reusable. All Ivy functions can be executed using TensorFlow, PyTorch, MXNet, JAX and NumPy, without any change to the code.

We are in talks with developers from Google, Facebook, NVIDIA, Hugging Face, and other top software companies who would like to use Ivy in their popular open-source projects, to instantly support all frameworks. A few examples are: Ray, FastAI, HF Transformers, PyG, Pyro and Kornia.

We've just raised a round of venture funding (YC, Abstraction Capital, Essence, A.Capital). We’re looking for talented developers to join this ambitious endeavor, just as we’re getting started.

About the role

Skills: Torch/PyTorch, Git, Python, GPU Programming, TensorFlow, Linux, Docker

We are looking for talented developers to join our team full-time, to extend Ivy’s codebase as we expand into the PyTorch Ecosystem and beyond. We are hiring worldwide. During the role, your tasks would include:

● Helping to manage and oversee the work of our ML research engineering interns

● Helping to design Ivy's graph compiler and transpiler, enabling automatic code conversions between frameworks

● Working alongside our open-source partners, helping to incorporate Ivy into their popular repositories

● Implementing SOTA models in Ivy, and adding these to our model hub

Requirements

● Very strong Python skills, with deep expertise in at least one of: PyTorch, TensorFlow, JAX

● Very strong skills in recursive programming. Check out the Ivy Container class

● A passion for Machine Learning research, and for our vision to unify the ML frameworks!

Technology

Ivy is 100% Python, with very few dependencies. Ivy does not have a dedicated low level backend (C++, CUDA, TorchScript etc.). Instead, Ivy wraps the functional APIs of existing ML frameworks. Design-wise, Ivy makes extensive use of: recursion, function decorators, property-based unit testing (hypothesis) and monkey-patching (dynamic code changes). Check out our open-source code at: https://github.com/unifyai/ivy

Interview Process

Apply Here: https://deep-ivy-ltd.breezy.hr/

Applications via YC are not monitored

a) paid open-source coding assessment, where you must complete one of the open tasks on the GitHub repository, and then create a PR to merge the completed task into master.

b) 10 recorded video responses (60 seconds each) to both high-level and technical questions about Ivy and it's place within ML.

c) 30-minute technical interview with myself, with some white-board coding exercises around some of the more technical aspects of Ivy's design, including the graph compiler and the transpiler.

d) potential offer!

Ivy remote jobs

ML Engineer intern

ML Engineer