Earthdaily Analytics

OUR VISION
At EarthDaily Analytics (EDA), we strive to build a more sustainable planet by creating innovative solutions that combine satellite imagery of the Earth, modern software engineering, machine learning, and cloud computing to solve the toughest challenges in agriculture, energy and mining, insurance and risk mitigation, wildfire and forest intelligence, carbon-capture verification and more. 
EDA’s signature Earth Observation mission, the EarthDaily Constellation (EDC), is currently under construction. EDC will be the most powerful global change detection and change monitoring system ever developed, capable of generating unprecedented predictive analytics and insights. It will combine with the EarthPipeline data processing system to provide unprecedented, scientific-grade data of the world every day, positioning EDA to meet the growing needs of diverse industries.
OUR TEAM 
Our global, enterprise-wide team represents a variety of business lines and is made up of business development, sales, marketing and support professionals, data scientists, software engineers, project managers and finance, HR, and IT professionals. Our Data & Platform team is nimble and collaborative, and in preparation for launching a frontier and disruptive product in EDC, we are currently looking for a Sr. Software Engineer (ML Researcher) to join our crew! This is a Vancouver-based hybrid position with 3-days per week in-office required.
PREPARE FOR IMPACT!
As a Sr. Software Engineer (ML Researcher), you will be a core contributor to the research, design, and implementation of EarthDaily’s large‑scale geospatial foundation model for agriculture. You will combine deep expertise in modern deep learning and foundation model architectures with hands‑on development on earth observation datasets to push the state of the art for geospatial foundation model technology, leveraging the EarthDaily Constellation’s unique temporal, spectral, and spatial characteristics.
KEY RESPONSIBILITIES:

• Research, design, and validate deep learning architectures for large‑scale multi‑modal geospatial foundation models (e.g. combining optical imagery with weather and other contextual data) and evaluate trade-offs between architectures
• Lead large-scale training and fine-tuning of foundation models on large EO datasets
• Collaborate with machine learning infrastructure engineers on the team to optimize distributed training and cloud resource usage
• Collaborate with machine learning engineers on the team to define metrics and experiments to benchmark foundation model performance
• Participate in sprint planning, sprint reviews, sprint demos, sprint retrospectives 
• Ensure technical documentation and systems are created, maintained and operational 
• Grow your skillsets and share your experiences with the team 

YOUR PAST MISSIONS

• Degree in Computer Science, Math, Physics, Engineering, Geography, GIS or equivalent  
• Higher level education in machine learning, data science, remote sensing, or related field is an asset.
• 7+ years of combined software engineering and/or applied deep learning research experience, including geospatial foundation model research experience
• Proven experience designing and training algorithmically complex deep learning models for large scale datasets including earth observation data (e.g. Sentinel 2, Landsat)
• Hands on experience with modern deep learning architectures (e.g. CNNs, transformers, spatio temporal models), including understanding of trade-offs and how to adapt and combine architectural elements
• Experience working in cloud environments (e.g. AWS) for large scale distributed model training and data preprocessing
• Experience with Agile development, SCRUM and CICD processes, and collaborating with cross-functional teams
• Equivalent combination of education is accepted

YOUR TOOLKIT

• Excellent algorithmic, analytic, problem solving, debugging, optimization and code reviewing skills
• Physics and/or math knowledge an asset
• Good object-oriented and test-driven design skills 
• Good skills and knowledge of best practices in at least one programming language (e.g. Python, C++) 
• Proficiency in Python scientific stack and common tooling (e.g. NumPy, pandas, PyTorch, Jupyter).
• Familiarity with Python geospatial and EO tooling (e.g. GDAL, rasterio, xarray)
• Self-starter and self-learner attitude with the ability to manage and execute with minimal supervision 
• Ability to take initiative, commit and thrive in a fast-paced, deadline-driven environment 

OUR SPACE
We’d love to welcome you to our world of software for space. We have a shared passion for building production critical systems that generate near real-time views of Earth from satellites that power real-world applications like disaster mitigation, environmental monitoring and crop yield improvements. It’s a fun, fast paced, exciting  environment where we hold innovation, team work, honesty and trust as our core values.

To make the most innovative products that serve our customers, we recognize the role that each of us plays in Diversity and Inclusion at EarthDaily. We draw from our diverse crew of exceptional team members and encourage and empower our team members to express themselves regardless of identity, race, colour, ancestry, place of origin, religion, marital status, family status, physical or mental disability, sex, sexual orientation and gender identity or expression. 

YOUR COMPENSATION
Base Salary Range: $145,000-$170,000 CAD annually. This range is based on Vancouver, BC-derived compensation for this role and may differ for other geographies. The selected candidate's compensation will be determined based on multiple factors, including but not limited to job-related skills, experience, education, and location.
WHY EARTHDAILY ANALYTICS? 

• Competitive compensation and flexible time off 
• Be part of a meaningful mission in one of North America's most innovative space companies developing sustainable solutions for our planet
• Great work environment and team, with a waterfront head office location in Vancouver, BC. 

Powered by JazzHR

EOBQ4K7cne