Geospatial Data Engineer

Full time, London or UK-based.

About us:

BeZero Carbon is a global ratings agency for the Voluntary Carbon Market. Our ratings allow all market participants to price and manage risk. BeZero’s ratings and research tools support buyers, intermediaries, investors, and carbon project developers.

Founded in April 2020, our 150-strong team combines climatic and earth sciences, sell-side financial research, earth observation, machine learning, data and technology, engineering, and public policy expertise. We work from five continents.

www.bezerocarbon.com

Background on the role:

BeZero is looking for a mid- or senior-level geospatial data / back-end engineer to join our existing geospatial engineering & machine learning team. This team is part of our broader data organisation and is responsible for building geospatial data and machine learning products for our client-facing platform and internal teams. The team works closely with colleagues in our ratings, research, product, and technology teams.

You’ll be responsible for building geospatial data products and algorithms that directly affect the way our ratings and research teams analyse carbon offset projects. You’ll therefore work closely with researchers in our Geospatial Research team and ratings scientists in our Ratings team. We process large-scale satellite imagery data sets (think about any of the public NASA and ESA missions) of different types (optical imagery, radar, SAR) for most of our use-cases, but also leverage raster and vector data from partnerships we have with data vendors.

These are some of the projects members in our team have been working on recently:

• Building a scalable data pipeline that generates a cloud-free mosaic of a specified area based on Sentinel-2 imagery.
• Integrating fire event data from NASA for carbon offset projects in our client-facing platform.
• Making our raster data (cloud-optimised geotiffs) compliant with the STAC specification and deploying a STAC server so that our internal analysts can access all our geospatial data with ease.
• Developing our dynamic baselines machine-learning product that directly analyses the effectiveness of nature-based carbon offset projects.

If you’re excited by working on such problems and making impactful contributions to data in the climate space, then we’re looking for you.

Tech stack

As a data team, we have a bias towards shipping products, staying close to our internal and external customers, and end-to-end ownership of our infrastructure and deployments. This is a team that follows software engineering best practices closely. Our data stack includes the following technologies:

• AWS serves as our cloud infrastructure provider.

• Snowflake acts as our central data warehouse for tabular data. AWS S3 is used for any of our raster data, and we use PostGIS for storing and querying geospatial vector data.

• We use lots of technologies from the Python geospatial data stack: packages like gdal, rasterio, xarray, geopandas and tools like STAC and zarr.

• Our computational jobs are executed in Docker containers on AWS ECS, and we use Prefect as our workflow orchestration engine.

• AWS Sagemaker acts as a platform for data science and research teams to develop data pipelines and machine learning models. We use Weights & Biases for model experimentation and versioning.

• Metabase serves as a dashboarding solution for end-users.

• GitHub Actions for CI / CD.

We are a remote-friendly company and many of our colleagues work fully remote; however, for this position, we will only consider applications from candidates based in the UK. If you are located in or near London, you are welcome (but not required!) to work from our London office.

Responsibilities:

You will be an individual contributor in our geospatial machine learning team, focused on developing and maintaining geospatial data products to be deployed on our carbon markets platform or used internally by our ratings team.

You will build automated data pipelines that collect and manipulate large geospatial data sets (such as optical satellite imagery, SAR and LiDAR, climate data and others) and deploy them in our cloud infrastructure.

You will work with our internal research and ratings teams to integrate the outputs of (analytical) data pipelines into BeZero’s business processes.

You’ll be our ideal candidate if:

You care deeply about the climate and carbon markets and are excited by solutions for decarbonising our economy.

You are a highly collaborative individual who wants to solve problems that drive business value.

You have at least 4 years of experience building ELT/ETL pipelines in production for data engineering or machine learning use cases, using Python and SQL.

You have experience dealing with a variety of geospatial data formats (e.g., netCDF, cloud-optimised geotiff, geoJSONs, zarr) and geospatial SQL (PostGIS) and Python packages (e.g., xarray, rasterio, shapely, gdal).

You have hands-on experience with workflow orchestration tools (e.g., Prefect, Dagster, Airflow), containerization, a cloud platform (we use AWS but any cloud platform will do) and the Python scientific computing stack (NumPy, SciPy, matplotlib, pandas, etc).

You can write clean, maintainable, scalable, and robust code in Python and SQL, and familiar with collaborative coding best practices (e.g., for Python PEP8 code style, unit testing, continuous integration tools such as flake8, black, isort, etc).

You are well-versed in code version control and have experience working in team setups on production code repositories.

Bonus points (but we’d still like to hear from you if you don’t have experience in any of these)

You have experience with machine learning platforms (like AWS Sagemaker, MLFlow, Weights & Biases) to design, test, serve and monitor geospatial ML pipelines.

You have experience in deploying cloud resources using tools such as AWS Cloud Formation, Terraform, etc.

You have experience in productionising and analysing specific satellite imagery data types like SAR, LiDAR, or RADAR or another remote sensing domain.

Research has shown that women are less likely than men to apply for a role if they don’t have experience in 100% of the requirements outlined in a job description. Please know that even if you don’t have experience in all the areas above but think you could do a great job and are excited about shaping company culture, finding great people, and building great teams, we’d love to hear from you!

What we’ll offer:

• Competitive renumeration and opportunity for equity in a rapidly growing VC-backed start-up through share options, plus comprehensive benefits package.
• Ability to learn and develop alongside a range of sector specialists from the scientific, technology, economic and business community.
• Opportunity to be part of a fast-growing data organisation that is central to the company, with clear opportunities to set the pace and vision for the team and to either develop in the future as an IC, into a deep tech expert in carbon markets, or as an engineering manager, leading the day-to-day of the team.
• Flexible working arrangements: A central London office space (Hoxton) if you wish to work from an office, but opportunity to work completely remote as well.

Our interview process:

• Initial screening interview with recruiter (15 mins)
• Introduction call with Chief Data Strategist (30 mins)
• 2x Technical interview with members from the data engineering & science team (60-90 mins)
• Reference checks + offer

We value diversity at BeZero Carbon. We need a team that brings different perspectives and backgrounds together to build the tools needed to make the voluntary carbon market transparent. We’re therefore committed to not discriminate based on race, religion, colour, national origin, sex, sexual orientation, gender identity, marital status, veteran status, age, or disability.