Deep Learning Algorithm Researcher

Company Overview

Edgify is a revolutionary VC-backed startup, based in the Bursa area in Ramat Gan, aiming to change the way Deep Learning models are trained. We have developed a framework that allows training models on edge devices, where the data is generated, in a collaborative fashion, creating one robust model that was trained on all the data distributed between the devices, without the data actually leaving the edge. Our technology allows accessing data previously unavailable for privacy reasons and reducing data transfer and computation costs by leveraging the edge devices computation power. Our clients come from a variety of industries: Retail, medicine, manufacturing, connected cars etc.

Job Description

Our algorithm team deals with a variety of challenges associated with training DL models on the edge. We believe in “diving deep” into our solutions and many of our discussions end up in scribbling equations on the board.

Some of the cool research topics we’re working on:  Federated Learning, Model Compression, Online Learning, NON-IID data distributions, Differential Privacy etc.

As a researcher on the algorithm team, you will be given complicated problems to research and will be expected to lead the research process from the phase of literature review right up until the integration of the solution into the system. Our algorithm framework sits at the core of the Edgify product, and as part of the job you’ll be expected to participate in the development and improvement of the framework.

Requirements

  • MSc / PhD in Computer-Science, EE, Physics, Mathematics (or similar).
  • 2+ years of experience in ML/DL/CV development. Academic experience may be considered only in the case of hands-on research in DL with a relevant framework (see below).
  • Strong Python development skills.
  • Familiarity with one of the major DL frameworks: PyTorch /TensorFlow / MXNet.
  • Advantage: C++ development skills.

To apply for this job please visit edgify.bamboohr.com.

פרסם משרה
X