Назад
Company hidden
10 дней назад

Data Scientist Lead (AI/HPC)

Формат работы
onsite
Тип работы
fulltime
Грейд
lead
Английский
c1
Страна
Brazil
Вакансия из списка Hirify.GlobalВакансия из Hirify Global, списка международных tech-компаний
Для мэтча и отклика нужен Plus

Мэтч & Сопровод

Для мэтча с этой вакансией нужен Plus

Описание вакансии

Текст:
/
TL;DR
Data Scientist Lead (AI/HPC): Leading the development and productionization of AI and data technologies for seismic processing, imaging, and interpretation across GPU/CPU HPC clusters and AWS with an accent on deep learning, scientific software, and scalable data platforms. Focus on designing distributed training and inference, optimizing CUDA/MPI workloads, and turning research prototypes into tested, deployable production software.

Location: Presential, based in Rio de Janeiro, Brazil, with Brazilian residency and work authorization; occasional travel required

Company

hirify.global develops AI, data, scientific software, and energy data technologies, including seismic processing solutions, the MDIO open-source data format, and the hirify.global Data Verse platform.

What you will do

  • Lead, mentor, and grow data scientists and scientific software developers while setting technical goals, coding standards, and development paths.
  • Define architectures for ML models, data pipelines, and HPC/cloud workflows in collaboration with the hirify.global Data Science team in Houston.
  • Oversee deep learning for seismic processing, imaging, and interpretation, including seismic foundation models, ViT, and 3D CNNs such as SaltNet.
  • Profile, parallelize, and optimize training and inference across GPU/CPU clusters, CUDA, MPI, multi-node environments, and AWS.
  • Drive MDIO, cloud-native data practices, and integrations with hirify.global Data Verse, OSDU, and Prediktor operational data.
  • Move research prototypes into production through packaging, testing, CI/CD, model registries, deployment, user support, open-source contributions, patents, and publications.

Requirements

  • Bachelor’s or Master’s degree in Geophysics, Physics, Computer Science, Applied Mathematics, Engineering, or a related field, plus 5+ years leading R&D or software projects.
  • 3+ years leading data science, research, or scientific software teams, with demonstrated delivery of software to production.
  • Expert Python with NumPy, SciPy, and PyTorch, plus modern C/C++; Fortran is a plus.
  • Strong experience with MPI, CUDA, OpenMP, or other parallel programming models on GPU and CPU HPC architectures.
  • Strong experience debugging, profiling, and optimizing scientific code using tools such as Nsight, perf, or VTune.
  • Experience with distributed deep learning training and inference, including DDP/FSDP and mixed precision, as well as MLOps practices covering experiment tracking, model registries, and deployment.
  • Experience with UNIX/Linux, POSIX programming, Git, automated testing, and CI/CD.
  • Experience with multidimensional and chunked data formats such as MDIO, Zarr, and SEG-Y; Dask/xarray; AWS S3 and EC2 GPU; Docker, Kubernetes; and HPC schedulers such as Slurm.
  • Experience with seismic processing, imaging, or interpretation algorithms. Fluent English and Portuguese, Brazilian residency, and Brazilian work authorization are required.

Nice to have

  • Fortran experience, publications or patents in ML, HPC, or geophysics, OSDU data platforms, master data management, industrial time-series data, OPC UA, and contributions to open-source scientific software such as MDIO.

Culture & Benefits

  • Work as one team with hirify.global Data Science in Houston using a shared codebase and toolset.
  • Collaborate with geophysicists, imaging, HPC/IT, and product teams across the energy data value chain.
  • Work in a presential R&D hub in Rio de Janeiro with occasional travel.

Будьте осторожны

Если работодатель просит войти в их систему, используя iCloud/Google, прислать код/пароль, запустить код/ПО, не делайте этого. Это мошенники. Обязательно жмите «Пожаловаться» или пишите в поддержку. Подробнее в гайде →