Senior Data Scientist

This Senior Data Scientist role is broad – we are looking for someone who can sit at the intersection of data engineering, ecological data science, and technical product ownership. As Senior Data Scientist, you will be responsible for day-to-day data management and the building of automated data processing pipelines and visualizations. You will also be involved in applying discriminative machine learning and ecological modeling approaches to processing and analyzing biodiversity data and designing and building ecological decision support tools within our Natural State Analytics platform. This role owns data pipeline design, requirements, testing, and validation. Pipeline implementation may be carried out by you, the Technology team, or collaboratively depending on the project. This will include designing dynamic, structured field survey forms (in ODK) and data ingestion pipelines that ensure raw field data are cleaned, processed, and joined correctly to become analysis-ready datasets that can be exported for different end users. Your role will also include analyzing field datasets into decision-support metrics that can be displayed on platform dashboards and be used to generate project reports, under guidance from Project Delivery and Biometrics. Our Project Delivery and Biometrics teams decide how data collection should be done, what quality checks matter, and what data dashboards need to show. You will turn these requirements into precise, buildable specifications for the Technology team to implement on the platform and then verify what gets built. You will not need to write all the backend code yourself, Natural State's Technology team owns most implementation, but you do need to be comfortable building data pipelines (in SQL) and writing and executing data analysis functions (in Python or similar). Most importantly, you need to understand the Natural State science database and principles deeply enough to specify exactly what should be built, ask the right questions before it's built, independently check the result once its built, and communicate that information clearly to external technical and non-technical audiences. There is a lot of scope for growth within this role. We envision the first 6-12 months to be heavily focused on creating and improving data management pipelines. Once those systems become less time consuming to maintain, we hope you will bring your imagination and expertise to help us build data tools within the Natural State Analytics platform that inform better decision making and help us achieve our mission of Restoring the Natural World.

This role is for you if…

You have a strong technical background and are wanting to use your skills on real-world, applied biodiversity conservation and restoration projects.

You have ideas about how biodiversity data can be leveraged to make better decisions and you are excited to implement these.

You are detail-oriented and diligent, with a keen understanding of the importance of well-curated data and a passion for creating pipelines to support this.

You thrive in a remote work environment and are able to effectively manage your own workload without needing too much top-down direction.

You enjoy working in a small, high-performance team and pitching in where you’re needed.

Requirements

Must have: 5+ years of work experience and a degree in data science, computer science, statistics, mathematics, quantitative ecology or a related field plus experience working with ecological/biodiversity data (e.g. camera trap images, passive acoustic monitoring recordings, vegetation surveys, animal surveys, species lists, soil carbon, biomass, remote sensing observation, climate etc.).

Must have: Strong Python for scientific data work – pandas or polars for data handling, plus the analysis stack (numpy, scipy, statsmodels, scikit-learn or equivalent). Comfortable writing validation scripts that catch schema mismatches early and explain clearly what broke.

Must have: Strong statistical reasoning, including an understanding of concepts such as confidence intervals, uncertainty estimation, GLMs, and discriminative machine learning models.

Must have: Confident in SQL, joins, CTEs, window functions.

Must have: Comfortable investigating data issues independently in pgAdmin, DBeaver, or similar.

Strongly desired: Able to handle spatial data in PostGIS, QGIS, and GeoPandas – raster algebra, coordinate reference systems and reprojection, vector versus raster, and the common ways location data breaks.

Strongly desired: Experience working with remote sensing data.

Nice to have: Experience with ODK, KoboToolbox, Survey123, or a comparable field data collection platform (ODK Central admin experience is a plus).