Data Scientist, Machine Learning

This is a fully remote position, open to applicants in Mexico.

📋 Description

• Develop and calibrate change-detection and anomaly models utilizing multi-temporal Sentinel-1/2 imagery across pipeline corridors.

• Acquire normal-terrain baselines per site and validate detections against ground-truth event logs by assessing detection rate, lead time, false-positive rate, and AUC.

• Optimize geospatial foundation models such as Prithvi-EO using LoRA/PEFT on a limited amount of labeled data.

• Apply SAR techniques like amplitude change, coherence, and pixel-offset tracking to evaluate pipe and dune movements.

• Create dune-migration tracking through optical flow or feature tracking, which includes assessing migration direction and mobility indices.

• Engineer data ingestion from Copernicus/CDSE, Sentinel Hub, and STAC, integrating optical, SAR, DEM, and ERA5 wind data.

• Design labeling strategies for encroachment masks and severity, while creating train/validation splits that prevent data leakage.

• Transparently communicate results and limitations to both technical and business stakeholders.

• Engage in comprehensive digital transformation projects for global leaders through Sequoia Connect's IT services partnership.

⛳️ Requirements

• Over 7 years of practical data science / ML experience with a focus on geospatial remote sensing.

• Proficient in Python programming, including libraries such as numpy, rasterio/GDAL, xarray, scikit-image, and geopandas/shapely.

• Familiar with optical and SAR data, including spectral indices, backscatter/dB, resolution trade-offs, and revisit times.

• Expertise in deep learning using PyTorch, encompassing transfer learning and LoRA/PEFT fine-tuning.

• Demonstrated capability in model validation and calibration, including ROC/AUC metrics, thresholding, cross-validation, and managing weak/few labels.

• Experience with time-series analysis and change-detection techniques.

• Knowledge of coordinate reference systems, such as UTM and reprojection methods.

• Strong resilience, emotional intelligence, and a commitment to agile delivery.

• Comprehensive understanding of the differences between coding and engineering.

• Proficient in oral English.

• Proficient in Spanish.

• Desired experience with InSAR / SAR offset tracking using SNAP, ISCE, or similar tools.

• Familiarity with geospatial foundation models like Prithvi-EO, TerraTorch, and HLS, as well as segmentation techniques is preferred.

• Knowledge of Copernicus/CDSE, Sentinel Hub, STAC, and Planetary Computer is beneficial.

• Exposure to aeolian geomorphology, dune dynamics, or the oil & gas / pipeline-integrity sectors is a plus.

• Desired experience with MLOps and cloud environments, including containerization, scheduled inference, and geospatial data pipelines.

• MSc/PhD in Remote Sensing, Geospatial Science, Earth Observation, CS/ML, Physics, or equivalent experience is preferred.

• Familiarity with cloud-native foundations or AI coding assistants is advantageous.

• Preference for candidates with experience in Space Tech, although it is not a requirement.

🏝️ Benefits

• Flexible remote work environment.

• Opportunities for global career advancement and engagement in high-impact projects.

• Involvement in comprehensive digital transformation initiatives for leading global organizations.

• Collaboration with international teams within a vast network of expertise.

• Recognized workplace as one of the most sustainable corporations globally.