Graph Infrastructure Engineer

Update: 2021-12-20 – these roles are filled, although we're looking for people with HPC experience in C++ Posted: 2021-08-18

Hiring for software engineering roles to support graph data science infrastructure, for use in industrial AI at scale.

Terms: Independent contractor, up to US$200/hour depending on experience, for an initial six month period. Must be available for some working sessions (in English) over video calls during CEST and US Pacific timezones.

This project is supporting a large EU firm (manufacturing) within an ongoing project, so there's a high likelihood of continuation after the initial contract period.

Requirements: Required competencies include: open source integration, software development in Python and C++ on Linux, team coordination through GitHub, plus some familiarity with cloud computing and distributed systems. We do not expect candidates to have experience in all of the technology areas listed below and will train on-the-job as needed. That said, some priority will be given to candidates with experience covering more of the listed skills.

Technology Areas: Overall, people on this team will be responsible for assisting an existing project which combines use of the following technologies:

• Python development, plus PyPi, Conda, pytest, MkDocs
• C++ development, plus Make, Boost, etc.
• Cython
• unit testing and performance analysis for C++ and Python on Linux
• high performance containers in C++
• GitHub, including CI with Actions
• Data Science
  • Pandas, Dask, sklearn, Jupyter, etc.
  • extras: Graphistry, NetworkX, PSL
• Data Engineering
  • Kubernetes, Terraform
  • extras: AsyncAPI, Kafka
• Azure cloud
• Arrow/Parquet
• Ray
  • extras: distributed design patterns, composable futures, performance trade-offs
• RAPIDS
  • extras: GraphBLAS, CUDA
• distributed systems architecture
  • extras: cloud computing cost optimization
• deep learning and graph embedding
  • extras: DGL, PyTorch, JAX

Big plus if you have prior experience working with distributed graph algorithms at scale.

Qualified for the role? To decide, first check these articles:

• https://medium.com/knowledge-technologies/graph-thinking-5e9c85134ab0
• https://www.nvidia.com/en-us/ai-data-science/resources/hardware-software-process-book/

Then check these open source projects:

• kglab
• ray_tutorial

If you're interested in the concepts described in the former and would be comfortable submitting pull requests on the latter then you are likely qualified to start.