we are looking for a Julia back-end developer and other related roles. You will be working on building the back-end components of a distributed machine learning system TIM (Tangent Information Modeller).
Your work will have an impact as TIM is being used in several industries (Energy, Retail, Finance, …) bringing automatic time-series model building to enterprises.
TIM implements InstantML - a modelling strategy that build high-quality models for time-series fast and automatic and is a next thing after AutoML. All the core micro-services are written in Julia to achieve high performance computing on several layers of abstraction.
Feel free to get in touch via a personal message or directly in this thread.
Can I ask why Julia? Why not something like Spark or Dask with Horovod?
Just wanted to tease out the pros and cons.
Well, what our product TIM does is (rather) orthogonal to e.g. Spark or Horovod / Deep Learning frameworks in general.
… and why Julia? Because it allows you to implement or the heavy lifting computation at ease and it is a full blown programming language that allows you to address also larger software projects. I had a talk on this in London at JuliaCon 2018 (JuliaCon 2018 | Our journey through the Perils of AOT compilation | Ján Dolinský - YouTube).