Data Scientist / Applied Scientist

Our Company

sustainacraft, Inc. focuses on nature-based solutions (NbS) for climate change and aims to maximize their effect through a more accurate carbon accounting system. While it is important to expand projects in forest conservation, sustainable forest management, and agriculture, it is equally imperative to ensure the transparency of such projects and the sound circulation of funds while increasing credibility.

We are currently working with environmental NGOs, financial institutions, and international development agencies in Southeast Asia, Latin America, and Europe to create a healthy financial cycle for nature conservation through two approaches: monitoring natural resources using satellite imagery and assessing appropriate reference levels using causal inference techniques.

Working Environment

Our Working Environment

We encourage asynchronous communication through services such as Notion, Slack, and Loom, making it possible for people to work regardless of their location. Although our day-to-day interactions are online and based on remote work, we provide an opportunity to work together in a relaxed environment near Tokyo once a week.

Our Colleagues

Members working with us come from diverse backgrounds, including consultants from professional consulting firms and foreign development banks, satellite image analysis specialists at satellite constellation companies, researchers in causal inference, and geographic information visualization specialists. Many of our members hold PhDs in engineering and economics.

Learning Opportunities

Our paper, presented in collaboration with the National Institute for Environmental Studies and Hitotsubashi University, won Best Paper at the Climate Change AI Workshop at NeurIPS 2022, a top-level conference on machine learning.

https://www.climatechange.ai/papers/neurips2022/55

Our strong R&D is a key component of our competitive edge.We also publish a regular newsletter on approaches and methodologies for forest and soil biomass estimation and project intervention effects.

https://sustainacraft.substack.com/

※ Regarding software engineering, we need to devise various ways to process satellite remote sensing information and causal inference technology at the operational level, and we are trying to do so through trial and error while also verifying cutting-edge technology.We would like to actively communicate engineering-related information, and we welcome anyone who is interested to spread such information through our newsletters and other means.In addition to our own engineers, we are supported by experienced outsourced engineers, and we believe that this is an ideal environment where we have the opportunity to absorb various points of views and knowledge, including from outside the company.

Job Description

Your responsibilities include:

• Research and development for forest carbon estimation from satellite imagery and other information sources, as well as algorithms for new products currently in progress.
• Design of field surveys for each project from a statistical point of view.
• Uncertainty assessment of estimation results based on field surveys and allometric equations.
• Design and development of web applications related to the above

We are looking for

• Being passionate about staying on top of cutting-edge technology trends on a daily basis and using those to create better products.
• Able to read scientific and technical papers routinely and critically, and reproduce and implement those papers.
• Able to express his/her knowledge in words and share with members on a daily basis.
• (Senior level or above) Able to take the initiative in designing product roadmap, taking into account trends in the carbon market.

Currently, our data scientists and applied scientists who design and develop algorithms implement web applications using Python-based frameworks.

The following requirements do not necessarily need to be fulfilled prior to application. For example, if you are a person who can cover the entire range of topics, or a person who has a deep knowledge of a specific topic, your career aspirations will be evaluated and salary will be determined accordingly. 

Requirements

Essential qualifications

• Understanding of the theory and practical experience in satellite image analysis (Experience in analyzing satellite imagery such as optical, SAR, LiDAR, hyperspectral, or similar areas)
• Theoretical understanding of statistics and computer vision
• Experience in data analytics and web application development using cloud services (GCP, AWS, etc.)
• English proficiency (At a minimum, you should be able to read English documents, especially scientific and technical papers. There will also be opportunities to write for English journals and discuss with overseas research institutions and researchers.)
• [for Applied Scientists only] Master's degree (preferably PhD) in the following (or similar) fields of expertise (computer vision, optical/microwave remote sensing, forest vegetation, environmental economics)

Preferred qualifications

• Experience in GHG emission estimation and observation of biodiversity in forests, peatlands and rice paddies.(Knowledge and research experience in the following topics:)
  • estimating above-ground biomass of forests and peatland condition
  • satellite missions and sensors (such as GEDI, PALSAR, MOLI, HISUI, GRACE)
  • crop models (such as DNDC)
• Experience in geospatial information processing (Experience with raster/vector data processing using GDAL, QGIS, ArcGIS, postGIS, etc.)
• Experience with Python-based web frameworks (Experience with Dash, etc.)
• Experience with distributed processing and high-performance computational processing using GPUs (Experience with RAPIDS, Dask, Vaex, etc.)
• Experience in stochastic programming and spatio-temporal modeling (Experience using Stan, PyMC3, Pyro, etc.)
• Experience in competitive programming such as AtCoder (At least AtCoder “light blue” equivalent)
• Experience with javascript-based visualization of geographic information

Selection Process

Application screening → First interview → Final interview
Please note that the number of screening may increase or decrease.