Join Black Swift Technologies and build the computer vision technology that lets purpose-built aircraft fly scientific payloads through the world's most extreme atmospheric environments, monitoring wildfires, volcanoes, tornadoes, and hurricanes to gather data otherwise out of reach.

Black Swift Technologies

Black Swift Technologies builds unmanned aircraft systems for the most demanding flight environments on Earth. Founded in 2011 in Boulder, Colorado, and born out of research flying into tornadic supercell thunderstorms, BST designs its aircraft, avionics, autonomy software, payloads, and ground control systems entirely in house. That level of vertical integration lets a small, senior team field capabilities most organizations only prototype.

At the core of every aircraft is SwiftCore, our flight management system developed from the ground up rather than assembled around an off the shelf autopilot. Owning the entire stack is why our systems survive conditions that ground everyone else, and why our engineers can innovate at any layer of the aircraft.

BST is now scaling from a research driven engineering firm into a production supplier of autonomous aircraft for environmental sensing and defense missions, expanding our platform family with the S3, building toward NDAA compliant sourcing and Blue UAS listing, and growing rapidly with the backing of our strategic partner KrateoSky.

What You Will Do

You'll build the onboard computer vision that turns raw imagery into mission useful information across Black Swift's aircraft and payloads, then get it running reliably on constrained edge hardware in the field, not just in offline metrics. You'll be hands-on from day one: writing model code, profiling inference pipelines, debugging sensor and integration issues, and pushing systems from prototype to fieldable runtime.

Responsibilities:

Design, train, evaluate, and refine computer vision models for onboard detection, tracking, classification, and scene interpretation
Build and maintain inference pipelines optimized for latency, throughput, memory, and thermal limits on edge accelerators like Jetson class devices, FPGAs, or NPUs
Work directly with autonomy, embedded, and payload engineers to integrate perception into the broader flight and mission software stack
Support bench, flight test, and post flight debugging to trace perception issues through the sensor and software stack and drive them to resolution
Develop repeatable evaluation workflows measured against real mission requirements rather than generic benchmarks

Who You Are

You've personally built, optimized, and shipped computer vision systems that ran on physical hardware in the field, not just in experimentation. You're equally comfortable in Python and C++, you use AI tools like Copilot or Cursor as part of your normal workflow to move faster, and you can trace a failure from sensor input through inference to downstream software and see it through to a fix.

Qualifications:

3-5 years of experience in computer vision, machine learning, robotics, autonomy, or a closely related domain
Fluency in Python and C++ with hands on use of PyTorch, ONNX Runtime, TensorRT, OpenCV, or comparable deployment tooling
Direct experience deploying models to embedded or edge hardware, including trade offs around compute, memory, and thermal behavior
Comfortable in terminal driven Linux/POSIX environments for deployment, profiling, and debugging on target systems
Eligible to obtain and maintain a U.S. DoD Secret security clearance
Preferred: experience with aerial, remote sensing, or geospatial imagery, visual navigation or visual inertial odometry, or defense adjacent environments like Blue UAS or NDAA sourcing

Benefits

$120,000 to $145,000 per year base salary, plus a comprehensive benefits package
100% paid healthcare benefits
401(k)
Flexible / Unlimited PTO
Free Eldora ski pass, plus flexible mornings on 6"+ powder days

Computer Vision Engineer