fixed-term project engagement｜Lead GeoAI & Data Science Engineer

[ ABOUT TENCHIJIN ] Tenchijin Inc. is a Tokyo-headquartered space-tech company and JAXA (Japan Aerospace Exploration Agency) certified venture. Through its land-evaluation and analytics platform, Tenchijin COMPASS, the company fuses earth-observation satellite data, geospatial and environmental layers, IoT and ground-sensor feeds, and clients' operational data with proprietary AI to deliver multimodal analytics for utilities and infrastructure enterprises. Tenchijin's flagship water-infrastructure solution, KnoWaterleak, assesses pipeline deterioration and leak risk from space-derived insights and is used by a growing number of water utilities to prioritize inspection and investment, reduce non-revenue water, and extend asset life. [ ABOUT THE PROJECT ] The Global South Project is an 18-month initiative (October 2026 – March 2028) to design, build, and deploy a scalable, secure, AI-ready analytics platform tailored to utilities and infrastructure operators across Global South markets. The project adapts Tenchijin's satellite-data and GeoAI capabilities to regions facing aging or rapidly expanding infrastructure, constrained budgets, and limited field-inspection capacity — delivering risk assessment and decision-support tools that improve infrastructure management workflows. All positions are contract-based and fully remote, operating as one distributed, cross-border team with English as the working language. [ ROLE SUMMARY ] This role sits at the core of Tenchijin's value proposition. You will develop the GeoAI machine-learning models that turn fused satellite imagery, geospatial layers, IoT feeds, and utility operational data into actionable risk-assessment and recommendation outputs — the same class of technology behind Tenchijin's land-evaluation and water-infrastructure risk products, adapted and extended for Global South infrastructure challenges. [ Reports To ] Senior Cloud Architect / Director of Product Management [ Project Language ] English (professional working proficiency or higher required) [ KEY RESPONSIBILITIES ] - Design, train, validate, and deploy machine-learning models for infrastructure risk assessment and recommendations (e.g., asset deterioration risk, environmental stress factors, prioritization scoring) using multimodal inputs. - Build geospatial feature-engineering pipelines that fuse satellite/earth-observation data (optical, SAR, thermal), terrain and environmental layers, IoT sensor streams, and client operational records. - Establish the ML lifecycle: experiment tracking, model registry, evaluation frameworks, retraining pipelines, and monitoring for drift in production. - Adapt models to data-sparse Global South contexts — transfer learning, handling incomplete asset records, and calibrating outputs against local ground truth. - Collaborate with the Lead Backend Engineer to productionize models: batch scoring, low-latency serving, and integration into the analytics APIs. - Define data-quality standards and validation for incoming geospatial and operational datasets. - Communicate model behavior, accuracy, and limitations to product leadership and client stakeholders in clear, non-technical terms. - Mentor engineers on geospatial data science practices and review analytical methodology across the project. [ WHAT WE OFFER ] - A central role in applying satellite data and AI to real infrastructure challenges, with measurable social and environmental impact in Global South markets. - Fully remote, cross-border collaboration with senior specialists across cloud, GeoAI, product, and design. - Competitive contractor compensation commensurate with experience and scope. - Direct exposure to earth-observation technology, including data ecosystems built with Japan's space agency (JAXA). Tenchijin Inc. is an equal-opportunity organization. We evaluate all applicants on qualifications and merit, without regard to nationality, race, religion, gender, age, or disability. TENCHIJIN INC. · GLOBAL SOUTH PROJECT

- 6+ years in data science / machine-learning engineering, with substantial production experience building, validating, and deploying ML models. - Strong proficiency with ML frameworks (PyTorch/TensorFlow, scikit-learn, XGBoost) and the Python data-science stack. - Experience building end-to-end data pipelines for large, heterogeneous datasets, including imagery, IoT, and time-series data. - Demonstrated experience taking models from experimentation to production: serving, monitoring, drift detection, and retraining. - Solid statistical grounding: model validation, uncertainty quantification, and communicating confidence and limitations honestly. - Professional working proficiency in English.

- Geospatial / remote-sensing expertise — tooling such as GDAL, rasterio, GeoPandas, PostGIS, or Google Earth Engine, and experience integrating satellite imagery (optical and/or SAR) into predictive models. Candidates coming directly from earth-observation organizations (e.g., Atlas AI, Google Earth Engine, NASA, JAXA, NOAA, ESA, or comparable) are especially encouraged to apply. - Domain experience in infrastructure, utilities, water, energy, or climate-risk analytics. - Experience with MLOps tooling (MLflow, Kubeflow, SageMaker, Vertex AI). - Publications, competition results, or open-source contributions in GeoAI / remote sensing. - Experience working in a cross-functional POD / squad delivery model.