Machine Learning / Computer Vision Engineer

SkyTruth

Shepherdstown, West Virginia

Job Type Paid permanent
Salary Details TBA
Deadline Jul 30, 2019

Are you a data and tech nerd? Are you passionate about big data, machine learning, and satellite imagery? Do you want to apply your time, energy, and skills to saving the environment, stopping pollution, protecting biodiversity, and improving the wellbeing of people around the world?

Then this position may be what you’re looking for.
And you may be the person we’re looking for.

SkyTruth is looking for a creative, innovative, experienced Computer Vision Engineer to join our growing team of committed tech + conservation advocates. This is an exciting opportunity combining cutting-edge earth observation data, cloud computing, and scalable analytics to solve the world’s environmental challenges.

In collaboration with our staff and our conservation, data, and technology partners, you will develop environmental change indicators — derived from near real-time satellite imagery and other remote sensing data — that enable intervention and other positive conservation actions and impacts. You’ll help save endangered species and their habitat by detecting new road construction encroaching on protected areas, and by measuring forest thinning and unauthorized land-clearing that reveals illegal logging or settlement. You’ll protect water supplies by identifying ground-water fed agriculture depleting aquifers, and mining and fracking operations that consume huge quantities of water. You’ll join a team that’s stopping pollution in the ocean by identifying ships that are dumping bilgewater, and by helping us shine a spotlight on oil leaks and spills from offshore platforms and pipelines.

What you’ll do

  • Design and implement highly scalable computer vision algorithms and image processing pipelines
  • Rapidly iterate algorithms to continuously improve precision and recall under a variety of conditions
  • Collaborate with SkyTruth staff and end-users to understand computer vision needs
  • Design and implement test methodologies for testing and accuracy assessment
  • Support build deployment, management, and monitoring systems
  • Provide production support for applications and infrastructure

Gotta haves

  • A desire to save the planet’s biodiversity, ensure that high conservation value areas are protected on land and at sea, and improve the lives and wellbeing of people everywhere
  • Comfort working with large, messy datasets that have sparsity in some dimension (e.g., spatially, temporally)
  • Comfort working with a small, dynamic team, with a willingness to jump in and help on other projects and gain experience in other areas
  • Experience working with relational databases (e.g., mySQL, PostgreSQL) and spatial extensions (e.g., PostGIS)
  • Excellent coding skills in Python, including deep familiarity with NumPy, SciPy, pandas, OpenCV, TensorFlow or PyTorch, and Keras
  • Expertise in traditional machine learning/modeling techniques like SVMs, K-nearest neighbor and Random Forest
  • Familiarity with deep learning, including semantic segmentation and object detection
  • Understanding and experience with Cloud architectures specifically: Google Cloud Platform (GCP), Microsoft Azure or Amazon Web Services (AWS)
  • Proven ability to work with software engineering / data science teams to develop and then deploy predictive models in a production environment
  • Experience with collaboration tools software like GitHub / GitLabs
  • An ability to communicate complex data science concepts and results in plain English to non-technical audiences
  • Deep experience and subject matter expertise with the analysis of satellite and/or aerial imagery using QGIS, Google Earth Engine, GDAL and other geospatial tools
  • Advanced degree and significant work experience

Nice-to-haves

  • Expertise in deep learning architectures including FCNs, ResNets, Faster R-CNN, and YOLT
  • Familiarity with DevOps tools like Docker and Kubernetes
  • Understanding of continuous delivery and deployment patterns and tools (Jenkins, Artifactory, Maven, etc.)
  • Experience designing and executing complex machine learning projects in a fast paced and agile development environment using the latest software development technologies, infrastructure, and methods
  • Experience working with distributed teams

Nimble, scrappy, and with a relatively flat hierarchy, SkyTruth encourages initiative and leadership at all levels. We innovate in the application of technology to demonstrate the art of the possible for conservation: opening new doors for scientific research and management, and creating new paths for sustainability, as exemplified by our Global Fishing Watch project in partnership with Oceana and Google. Our toolkit includes Google’s Earth Engine platform and cloud-compute suite, as well as data and analysis platforms operated by others. We use satellite imagery from the European Space Agency’s Copernicus program (Sentinel-1 radar for offshore pollution monitoring; Sentinel-2 multispectral imagery for terrestrial monitoring and analysis), as well as Landsat multispectral imagery and images from commercial vendors (Planet, DigitalGlobe, Airbus, MDA and others).

We cultivate a flexible work environment and a humane culture to foster work-life balance. Remote work is possible — we use Slack, Zoom and Google Meet/Hangouts.

HOW TO APPLY

By July 30, please fill out this form and upload a cover letter and resume that includes links to a few samples of your work (published writing, presentations, GitHub repos, etc.).

As an equal-opportunity employer SkyTruth welcomes applications from diverse candidates. We evaluate contractors, employees and applicants for employment based on merit, competence, qualifications and performance, and do not tolerate discrimination.

Location

213 West Washington Street, PO Box 3283, Shepherdstown, WV 25443-3283

When you apply, please indicate that you are responding to the posting on Conservation Job Board.