Unum AI
Senior DevOps Engineer: Databases & AI Startup
Unum AI, San Francisco, California, United States, 94199
Unum is the deep-tech startup reinventing Data-Lakes for extreme scale and AI!
You can think of it as Snowflake and OpenAI combined.
We are searching for a passionate and competitive Product Manager to join us in designing next-generation data infrastructure to empower million data-intensive and AI applications!
TasksYou will orchestrate software development and hardware management in one of the most advanced modern software projects. You will discover things about software and hardware that you won't learn anywhere else.
What kind of hardware, you ask?
Liquid-cooled state-of-the-art servers, with 64 core CPUs, terabytes of RAM, 200 GBit peer-to-peer mesh Mellanox networking. This also includes quad-GPU instances with Ampere, Turing, and Volta Nvidia GPUs. We use all of that for R&D and CI. The deployments are done across public clouds, including x86 and Arm CPUs, and specialized accelerators.
What kind of software products are we building?
We are building the fastest universal Database Management Software. It has both an open-source and closed-source part, split between GitHub and GitLab. The latter has more CI/CD around it, including a custom build system, combining Conan, PIP, Conda, and systems default package manager. It also comes with auto-generated Doxygen documentation, automated testing, test-coverage reports, and even automated continuous benchmarking via Intel oneAPI.
It has bindings for high-level languages like Python, Java, and GoLang. All of those bindings are tested and packaged separately. Similarly, Docker images and binaries must be generated for multiple operating systems, mainly focusing on Linux flavors: Ubuntu, CentOS, and Rocky Linux.
RequirementsResponsibilities
Reorganizing our build and CI/CD pipelines for mostly C++ and CUDA projects.
Packaging our products for a large matrix of Operating Systems and Versions.
Managing client data migrations in stateful cloud services.
Extend our continuous monitoring from internal to external infrastructure.
Qualifications
Python and Bash for scripting.
Kubernetes and Docker.
GitHub and GitLab pipelines.
AWS and GCP deployments.
#J-18808-Ljbffr
You can think of it as Snowflake and OpenAI combined.
We are searching for a passionate and competitive Product Manager to join us in designing next-generation data infrastructure to empower million data-intensive and AI applications!
TasksYou will orchestrate software development and hardware management in one of the most advanced modern software projects. You will discover things about software and hardware that you won't learn anywhere else.
What kind of hardware, you ask?
Liquid-cooled state-of-the-art servers, with 64 core CPUs, terabytes of RAM, 200 GBit peer-to-peer mesh Mellanox networking. This also includes quad-GPU instances with Ampere, Turing, and Volta Nvidia GPUs. We use all of that for R&D and CI. The deployments are done across public clouds, including x86 and Arm CPUs, and specialized accelerators.
What kind of software products are we building?
We are building the fastest universal Database Management Software. It has both an open-source and closed-source part, split between GitHub and GitLab. The latter has more CI/CD around it, including a custom build system, combining Conan, PIP, Conda, and systems default package manager. It also comes with auto-generated Doxygen documentation, automated testing, test-coverage reports, and even automated continuous benchmarking via Intel oneAPI.
It has bindings for high-level languages like Python, Java, and GoLang. All of those bindings are tested and packaged separately. Similarly, Docker images and binaries must be generated for multiple operating systems, mainly focusing on Linux flavors: Ubuntu, CentOS, and Rocky Linux.
RequirementsResponsibilities
Reorganizing our build and CI/CD pipelines for mostly C++ and CUDA projects.
Packaging our products for a large matrix of Operating Systems and Versions.
Managing client data migrations in stateful cloud services.
Extend our continuous monitoring from internal to external infrastructure.
Qualifications
Python and Bash for scripting.
Kubernetes and Docker.
GitHub and GitLab pipelines.
AWS and GCP deployments.
#J-18808-Ljbffr