Perennial is building the world's leading verification platform for soil-based carbon removal. The company is using advanced remote measurement technology, machine learning, ground observations, and satellite data to map soil carbon and land-based GHG emissions at continent-level scales to help the food supply chain decarbonize.
Requirements
- Strong proficiency in Python for data science (e.g. pandas, scikit-learn, xarray, numpy)
- Experience building machine learning, statistical, or time series models informed by remotely-sensed data or large spatial datasets
- Experience working in the soil carbon MRV space and familiarity with relevant methodologies and tools (e.g. VM0042, VM0032, VMD0053)
- Expertise in the open source geospatial python stack. Basic raster and vector operations, e.g. resampling, tiling, clipping, extracting, spatial statistics, harmonizing data
- Experience quantifying uncertainty of spatial maps, or more generally geostatistics or spatial stats, esp. with machine learning
- Experience working with Google Earth Engine and GCS
Responsibilities
- Build, improve, and deploy machine learning models for predicting soil carbon stock with remotely-sensed covariate data and limited training data
- End-to-end deliveries for our customers: train models, run predictions, and ensure quality results are delivered in customer reports
- Work with other data scientists, engineers, and policy experts to ensure that our data and methods comply with various standards and methodology requirements specific to a given project
- Characterize the accuracy and uncertainty of model predictions and demonstrate the dependence of performance metrics on the surrounding context and parameters of carbon projects
- Execute efficiently throughout full development cycle, from performing exploratory data analysis and initial R&D to rapid prototyping and hardening models for production
- Communicate your research internally and externally through detailed documentation, conference presentations, and peer-reviewed publications
Other
- Master's degree or Ph.D. in statistics, math, computer science, remote sensing, AI/ML, ecosystem science, soil science, geography, or a related STEM field
- 3–6 years of industry or research experience in data science, applied ML, geospatial analysis, or related fields
- Good communication and collaboration skills with functional and cross-functional teams
- Ability to independently manage a project and deliver results
- Startup experience or a strong entrepreneurial mindset (generally private company experience)
