Taichi Takayama
Geospatial Software Engineer
MOUNTAIN VIEW,CA
Summary
Experienced Geospatial Backend Engineer and Machine Learning Specialist
Proven expertise in developing scalable backend systems and processing geospatial data for advanced analytics and decision support. Extensive experience in integrating satellite, airborne, and ground sensor data for agricultural and environmental monitoring. Specializes in building geospatial pipelines, developing classification and predictive models, and implementing machine learning solutions for hyperspectral and time-series data. Adept at designing end-to-end systems, from preprocessing large datasets using PostGIS to deploying robust backend architectures for real-world services. Demonstrated leadership in driving cross-functional collaboration and delivering innovative solutions for complex geospatial challenges.
Overview
16
16
years of professional experience
Work History
Senior Software Engineering Manager and Project Le
Andes Ag, Inc.
03.2024 - Current
- Performed coding contributions at the same level as an individual contributor due to the small team size.
- Designed and implemented scalable backend pipelines for geospatial data processing, including Sentinel-2 image acquisition and cloud removal using multi-day image composites per month.
- Built systems to upload processed satellite images to Amazon S3 with STAC catalog management.
- Build and maintain four backend systems with GeoDjango for the following usages (1. frontend apps, 2. serving customized data originally from USGS, USDA, NASA and ESA, 3. backend health monitoring with Prefect, 4. MLOps server (MLFlow))
- Applied machine learning models for classification and segmentation tasks to enhance agricultural monitoring capabilities.
- Developed CI/CD pipelines for streamlined geospatial software deployments with AWS managed by Terraform.
- Collaborated with cross-functional teams, including scientists and product managers, to align technical goals with broader company objectives.
Engineering Lead
Andes Ag, Inc.
03.2022 - 03.2024
- Performed coding contributions at the same level as an individual contributor due to the small team size.
- Architected and maintained high-performance APIs and services for geospatial data analysis and querying.
- Designed database schemas optimized for managing satellite and hyperspectral data, including PostGIS integration for efficient geospatial operations.
- Led the design of backend systems for geospatial data ingestion, processing, and analytics pipelines.
Senior Remote Sensing Engineer
Andes Ag, Inc.
02.2021 - 02.2022
- Implemented distributed processing pipelines for satellite image analysis with a focus on scalability and accuracy.
- Developed prediction model of soil characteristics with satellite images (Sentinel 2, WorldView3, MODIS, VIIRS)
Senior Software Engineer
FarmX
09.2019 - 02.2021
- Designed and implemented scalable backend systems for processing and analyzing agricultural data using geospatial information.
- Built models for crop production prediction (vine yield) and anomaly detection, applying advanced machine learning techniques for time-series data.
- Preprocessed large geospatial datasets, including eliminating cloud and shadow pixels and applying interpolation methods, ensuring high-quality inputs for machine learning models.
- Developed efficient pipelines for integrating satellite data (MODIS, Landsat 8, Sentinel-2) with ground sensor data for real-time monitoring and yield estimation.
- Leveraged hierarchical Bayesian models and LSTM-based frameworks for predictive analytics.
- Key Contributions:1. Automated cloud-based workflows for data ingestion and processing using Python and PostGIS. 2. Enhanced model accuracy through statistical and spatial data analysis.
Lead Software Engineer / Co-Founder
Tenchijin
06.2019 - 12.2020
- Spearheaded the development of geospatial data pipelines and backend systems, leveraging PostGIS databases and Docker-based environments.
- Built analytical platforms for spatial and temporal data, including precipitation forecasting using deep learning (U-Net) and flight density maps for aviation.
- Conducted large-scale geospatial data analysis, integrating real-time and historical datasets for decision support systems.
- Delivered time-series analysis using Python and statistical methodologies to uncover patterns and enhance forecasting accuracy.
Software Engineer
AI/ML Startup (Stealth Mode) • StartX Accelerator,
01.2019 - 03.2019
- Developed semantic segmentation software for real-time video processing on NVIDIA TX2 for self-driving vehicles.
Researcher
Mitsubishi Research Institute, Inc.
04.2009 - 09.2018
- Planned and conducted airborne hyperspectral measurements in Indonesia for paddy and forest areas.
- Built prediction models for CO2 emission estimation using Landsat and MODIS data.
- Developed classification models for crop yield, biomass prediction, forest type classification, and palm tree disease detection.
- Conducted field measurements in tropical forests to collect ground truth data at Kalimantan peatland forests.
Visiting Scholar
Stanford University
06.2017 - 08.2018
- Researched state-of-the-art methodologies for image processing and design thinking in research and enterprise.
Education
Ph.D. - Engineering (Remote Sensing/Machine Learning)
The University of Tokyo
Tokyo, Japan 05.2001 -
Master of Science - Engineering (Remote Sensing)
The University of Tokyo
Tokyo, Japan 05.2001 -
Bachelor of Science - Engineering (Remote Sensing)
The University of Tokyo
Tokyo, Japan 05.2001 -
Skills
Distributed Systems: Scalable API development, high-performance data pipelines, cloud-native architectures with Django RestAPI, AWS ECS, ALB, S3 managed by Terraform
Backend Development: Python, Django, PostGIS, Docker
Geospatial Data Processing: Python (GDAL, Rasterio, xarray, PySTAC, GeoPandas, Shapely, GeoDjango, etc)
Cloud Platforms: AWS, CI/CD pipelines (AWS CodePipeline), data storage optimization
Machine Learning: PyTorch, Scikit-learn, lightgbm, tslearn, pymc, pyro, numpyro, monitored by MLOps (MLflow)
Collaboration: Cross-functional team leadership, technical alignment with scientific and product teams
Remote Sensing: Hyperspectral and multispectral analysis, preprocessing, sensor calibration
Publications
T. Takayama and A. Iwasaki, “Optimal wavelength selection on hyperspectral data with fused lasso for biomass estimation of tropical rain forest,” ISPRS Annals of Photogrammetry, Remote Sensing and Spatial Information Sciences, pp. 101–108, 2016. (Cited by 8)
T. Takayama, A. Iwasaki and O. Kashimura, “Optimal segmentation of classification and prediction maps for monitoring forest condition with spectral and spatial information from hyperspectral data,” 2014 IEEE Geoscience and Remote Sensing Symposium, Quebec City, QC, 2014, pp. 3498-3501. (Cited by 5)
T. Takayama, T. Ohki, H. Sekine, S. Ohnishi, S. Shiodera, M. Evri, and M. Osak, “Application of hyperspectral data for assessing peatland forest condition with spectral and texture classification,” 2013 IEEE International Geoscience and Remote Sensing Symposium - IGARSS, Melbourne, VIC, 2013, pp. 1007-1010. (Cited by 4)
T. Takayama, et. al., “Validation of BiPLS for improving yield estimation of rice paddy from hyperspectral data in West Java, Indonesia,” 2012 IEEE International Geoscience and Remote Sensing Symposium - IGARSS, Munich, 2012, pp. 6581-6584. (Cited by 3)
T. Takayama, T. Ohki, and T. Takeda, “Discrimination of peat swamp forest types with hyperspectral data,” 2015 7th Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing (WHISPERS), Tokyo, 2015. (Cited by 2)
T. Takayama, A. Iwasaki, Y. Yokota, T. Morota, J. Haruyama, T. Matsunaga, and M. Ohtake, “Validation of frame- transfer correction of SELENE/LISM/MI," in IEEE Transactions on Geoscience and Remote Sensing, vol. 49, no. 8, pp. 2911-2917, Aug. 2011. (Cited by 2)
T. Takayama, N. Yokoya, and A. Iwasaki, “Optimal hyperspectral classification for paddy field with semisupervised self-learning,” 2015 7th Workshop on Hyperspectral Image and Signal Processing: Evolution in Remote Sensing (WHISPERS), Tokyo, 2015. (Cited by 1)
K. Yoshida, T. Takayama, K. Fukuhara, A. Uchida, H. Sekine, O. Kashimura, "A Sparse Regularization Approach to Hyperspectral Image Analysis: An Application forRice Growth Monitoring and Yield Prediction in
Indonesia," Journal of The Remote Sensing Society of Japan, vol 32, Issue 5, pp. 287-299, 2012 (Cited by 1)
A. Uchida, H. Sekine, K. Fukuhara , K. Yoshida, T. Takayama, C Kobayashi, O. Kashimura, M. Evri, A. Wibowo, M. Sadly, Arief D.3., N. Pudi, S. Muljono, "Development of Hyperspectral Data Utilization Technology by Using Data Mining Method for Paddy, West Java, Indonesia" in Proceedings of the 30th Asian Association on Remote Sensing (AARS), pp. TS36–6., 2010 (Cited by 1)
T. Takayama and A. Iwasaki, “Selection of additional training data for improving accuracy of forest type classification using hyperspectral data,” 2016 IEEE International Geoscience and Remote Sensing Symposium (IGARSS), Beijing, 2016, pp. 3500-3503.
T. Takayama, H. Tsuji and H. Nagayama, “Evaluation for millimeter wave broadband wireless direct communication between airplane and ground,” 2014 International Symposium on Wireless Personal Multimedia Communications (WPMC), Sydney, NSW, 2014, pp. 734-739.
T. Takayama, H. Tsuji, and H. Nagayama, “Performance Evaluation of Radio-wave Monitoring with an Unmanned Aircraft System,” TJMW 2014.
T. Takayama, A. Iwasaki, J. Haruyama, T. Matsunaga, and M. Ohtake, “Generalized formulation of image correction applied to SELENE/LISM/MI,” International Symposium on Space Technology, Tokyo, 2008.
T. Ohki, K. Yoshida, H. Sekine, T. Takayama, T. Takeda, K. Hirose, M. Evri, M. Osaki, "Hyperspectral data application for peat forest monitoring in Central Kalimantan, Indonesia," International Society for Optics and Photonics, vol 8527, P85270J, 2012
Timeline
Senior Software Engineering Manager and Project Le
Andes Ag, Inc.
03.2024 - Current
Engineering Lead
Andes Ag, Inc.
03.2022 - 03.2024
Senior Remote Sensing Engineer
Andes Ag, Inc.
02.2021 - 02.2022
Senior Software Engineer
FarmX
09.2019 - 02.2021
Lead Software Engineer / Co-Founder
Tenchijin
06.2019 - 12.2020
Software Engineer
AI/ML Startup (Stealth Mode) • StartX Accelerator,
01.2019 - 03.2019
Visiting Scholar
Stanford University
06.2017 - 08.2018
Researcher
Mitsubishi Research Institute, Inc.
04.2009 - 09.2018
Ph.D. - Engineering (Remote Sensing/Machine Learning)
The University of Tokyo
05.2001 -
Master of Science - Engineering (Remote Sensing)
The University of Tokyo
05.2001 -
Bachelor of Science - Engineering (Remote Sensing)
The University of Tokyo
05.2001 -
Similar Profiles
GENESIS TAKAGENESIS TAKA
Operations Manager at Andes Express S.A.COperations Manager at Andes Express S.A.C
SOFÍA MANTILLASOFÍA MANTILLA
English teacher at Universidad De Los Andes - AndarEnglish teacher at Universidad De Los Andes - Andar
Andrea Belén Casanova IntriagoAndrea Belén Casanova Intriago
Extensionista Rural at Fundación para la Conservación de los Andes Tropicales- FCATExtensionista Rural at Fundación para la Conservación de los Andes Tropicales- FCAT
Matthew ZelayaMatthew Zelaya
Facilities Coordinator at CBRE @ GoogleFacilities Coordinator at CBRE @ Google
LAN YAOLAN YAO
Senior Software Engineer at Confluent, Inc.Senior Software Engineer at Confluent, Inc.