ML Engineer

📋 Job Overview

Vibrant Planet is seeking a ML Engineer to build, adapt, and operationalize foundation model-based deep learning systems that estimate forest structure metrics from remotely sensed data. The role involves fine-tuning geospatial foundation models, integrating trained models into automated production pipelines, and maintaining data infrastructure. This position requires collaboration across interdisciplinary teams and contributes to scientific knowledge dissemination.

🏢 About VIBRANT PLANET

We are a team of leaders in fire science, applied science, forestry, policy, and tech who work with land managers, community risk managers, utilities, and insurers to drive action that lowers the risk of destructive wildfire. Our cloud-based, AI-driven platform modernizes land management planning, community risk assessment, and monitoring through scenario building, decision support, and treatment outcome detection. Our fire science subsidiary, Pyrologix, produces leading wildfire science and models that quantify wildfire hazard and risk, and the benefits of mitigation action. This science powers the core Vibrant Planet platform and supports our work across utilities, insurance, and other sectors. Vibrant Planet is backed by climate and resilience leaders including Grantham Foundation, Earthshot, Elemental Excelerator, Ecosystem Integrity Fund, Cisco, and Halogen Ventures.

🎯 The Role

Vibrant Planet harnesses data-driven science and cloud-based technology to help make communities and ecosystems more resilient in the face of global change. Our ML Engineering team sits at the intersection of machine learning, remote sensing, and forest ecology—building the models, pipelines, and data products that power our Land Tender decision-support platform. In this role, you will build, adapt, and operationalize foundation model-based deep learning systems that estimate forest structure metrics from remotely sensed data.

✅ Key Responsibilities

• Adapt and fine-tune custom or publicly available geospatial foundation models as backbone architectures for domain-specific deep neural network heads that estimate forest structure metrics (canopy height, biomass, basal area, etc.).
• Prepare, curate, and manage training datasets from remote sensing sources (Sentinel-2, Sentinel-1, Landsat, lidar, NAIP) and field plot inventories.
• Evaluate model performance using standard remote sensing accuracy metrics and field-based validation data.
• Contribute to experiment design, hyperparameter optimization, and ablation studies in coordination with the Technical Lead ML Engineer.
• Integrate trained ML models into Vibrant Planet’s automated geospatial data pipeline as containerized, orchestrated inference services.
• Build and maintain STAC (SpatioTemporal Asset Catalog) infrastructure for data discovery, cataloging, and access control of ML model inputs and outputs.
• Design and implement larger pipelines composed of many smaller DAGs (Airflow), ensuring idempotency, observability, and fault tolerance.
• Maintain and improve data ingestion, preprocessing, and quality control workflows for satellite imagery and ancillary datasets.
• Monitor pipeline health and model drift; implement alerting and automated retraining triggers as needed.
• Develop model cards for summarization of modeling methods and performance.
• Write and contribute to scientific manuscripts describing methods, validation results, and novel applications.
• Serve as a cross-team link between SciDev, Data Engineering, and Product—translating requirements, communicating constraints, and aligning priorities.
• Document pipelines, model architectures, and operational procedures in team knowledge bases.
• Participate in code reviews, architectural discussions, and sprint planning.

📌 Required Qualifications

• M.S. in Computer Science, Machine Learning, Remote Sensing, Data Science, Ecology, or a related quantitative field (or equivalent work experience).
• 3+ years of experience developing, training, and deploying deep learning models (PyTorch preferred).
• Strong Python proficiency including data science stack (NumPy, pandas, xarray, scikit-learn).
• 3+ years of experience with geospatial data processing (rasterio, GDAL, geopandas, shapely).
• Experience building and maintaining data pipelines with workflow orchestration tools (Airflow, Prefect, Dagster, or equivalent).
• Proficiency with Git, Git

⭐ Desirable Experience

No desirable experience mentioned.

🎁 Benefits

Competitive

🛂 Visa & Eligibility

No visa or eligibility information mentioned.