Machine Learning (Data) Scientist


About Spire

Spire Global is a space-to-cloud analytics company that owns and operates the largest multi-purpose constellation of satellites. Its proprietary data and algorithms provide the most advanced maritime, aviation, and weather tracking in the world. In addition to its constellation, Spire’s data infrastructure includes a global ground station network and 24/7 operations that provide real-time global coverage of every point on Earth.

Job Description:

Spire Global is seeking a Machine Learning (Data) Scientist to support its development of artificial intelligence, statistical and probabilistic weather forecast products. This position relies on the candidate’s prior technical/scientific domain expertise to support the development of weather-related advanced machine learning and statistical inference applications based upon surface observations, radio occultation (GNSS-RO), reflectometry (GNSS-R) and other types of satellite data. The successful candidate will join the Statistics and Machine Learning team under the weather division, at the Spire office in Luxembourg, and will work closely with team members in Spire’s Boulder, CO, USA, office.

Responsibilities of your role:

The candidate will work with the team to develop advanced Machine Learning (ML) and Artificial Intelligence (AI) based weather technologies for global and regional forecasting. The successful candidate will work closely with data assimilation, modeling, software engineer, and product teams’ members to implement and evaluate machine learning and statistical weather models and make use of large datasets of surface and Spire satellite observations. The applicant will be also responsible to deliver high quality (production ready) code following best software engineer practices. Collaboration in the following areas will be essential:

  • Propose and implement innovative machine learning, predictive modeling, and Bayesian inference approaches to increase the forecast skills of Spire weather forecasts/products.
  • Develop and implement state of the art weather post-processing and downscaling techniques.
  • Explore different standard and non-standard data sources including Spire GNSS-RO and GNSS-R measurements, feature selection and sensitivity techniques to identify predictors for weather data-based models.     
  • Working with meteorological data sets from various sources.
  • Develop software packages to ingest new sources of weather data into the machine learning and statistical inference models.
  • Develop automated verification and validation of the scientific software packages.
  • Working with the software engineering team to define most effective software solutions including transition to operations.
  • Presenting research findings at scientific conferences or workshops.

Basic qualifications:

  • Master of Science degree in Data Science, Computer Science, Applied Mathematics, or Atmospheric Science or equivalent.
  • Working experience with state-of-the-art machine learning models including deep learning, regression trees, linear/non-linear regression, and classification methods.
  • Working experience in applying machine learning
  • Experience working with Python, including packing/deployment (Conda, PyP, Anaconda), experience and exposure to the scientific Python stack (NumPy, Pandas), and advanced machine learning libraries (Scikit-Learn, Xgboost, Tensorflow, Keras, Pytorch, FastAI, Dask-ML).

Preferred qualifications:

  • PhD degree in Data Science, Applied Mathematics, Atmospheric Science or Computer Science or equivalent.
  • Working experience with data science pipelines.
  • Experience working with GNU/Linux.
  • Prior experience working with modern software engineering best practices: Agile methodologies, revision-control systems, testing & code quality tools (regression, unittest, Pylint), continuous integration (Bitbucket Pipelines).
  • Working experience with ensemble of machine learning models, advanced statistical inference and uncertainty quantification.
  • Working experience with weather post-processing Bayesian techniques.
  • Experience with numerical weather prediction, weather applications, and/or data formats common in the weather domain (BUFR, GRIB, NetCDF).
  • Strong Object-Oriented Programming skills.
  • Experience with distributed parallel programming systems, frameworks, libraries such as MPI4PY, DASK, Hadoop, Spark, etc.
  • Experience with compiled languages (C, C++) and scientific/numeric applications.
  • Experience with cloud platforms (AWS, Google Cloud, etc.).

Spire is Global and our success draws upon the diverse viewpoints, skills and experiences of our employees. We are proud to be an equal opportunity employer and are committed to equal employment opportunity regardless of race, color, ancestry, religion, sex, national origin, sexual orientation, age, marital status, disability, gender identity or veteran status.