Senior Computer Vision Engineer

Skyways designs, builds, and operates fully autonomous long-range cargo aircraft. Founded in 2017, the company has spent the last eight years developing and deploying autonomous logistics systems for real-world operations.

Our V2 aircraft carries 30 lbs up to 500 miles, while our next-generation V3 carries 100 lbs over 1,000+ miles with 20+ hours of endurance. Both use a hybrid-electric architecture that takes off like a helicopter and cruises like a plane, enabling long-range autonomous delivery without traditional runway infrastructure.

Today, Skyways aircraft operate across three continents in controlled national airspace under FAA oversight and in support of U.S. military operations. We are transitioning from prototype development to full-rate production and scaling toward large autonomous cargo fleets.

Backed by Y Combinator and a $37M AFWERX STRATFI award from the U.S. Air Force, Skyways is creating a new form of transportation to advance our civilization from Austin, Texas.

Skyways aircraft have to land themselves, every time, often in places they have never seen before. This role owns the perception that makes that possible.

You will own the onboard vision system end to end, across both landing modes: target-based precision landing, where the aircraft locks onto a known marker and nails its touchdown, and targetless landing, where there is no marker and the aircraft has to read the terrain itself, find a safe spot, and reject the bad ones. This is hard, high-stakes perception running in real time at the edge, and it directly decides whether a mission ends in a clean delivery or a failure.

You will set the technical direction for landing perception, make the architectural calls that hold up across V2 and V3, and turn real flight data into a system that gets more reliable with every sortie.

We are looking for someone who has already owned a perception domain on a real platform that moves through the world, who can define the roadmap rather than execute one, and who wants their work measured in successful landings, not benchmarks.

• Own the landing perception stack end to end: target detection and relative pose estimation for marker-based landing, and safe-zone selection, hazard, and slope detection for targetless landing
• Set the architecture and technical roadmap for landing perception across the V2 fleet and V3 flight test program
• Build real-time CV inference that runs within the compute, power, and latency budget of an embedded flight system
• Make the fusion decisions: where camera, depth, stereo, or LiDAR earn their place in the stack and where they do not
• Define the datasets, evaluation strategy, and regression tooling that keep landing performance honest across releases
• Turn flight data, telemetry, and field failures into model and pipeline improvements through a tight iteration loop, and build the loop itself
• Debug real-world perception failures and close the gap between simulation, ground testing, and live flight
• Represent perception in cross-functional decisions with autonomy, GNC, and flight operations, including how perception integrates into the full landing sequence and what it means for flight readiness
• Raise the bar for the engineers around you through design review, mentorship, and the standards you set in the codebase

• Deep, hands-on ownership of perception on autonomous landing or a closely adjacent problem: target detection and relative pose, or safe-zone and hazard selection from terrain and depth.
• 6+ years post-grad shipping production CV software, with strong CS fundamentals across algorithms, data structures, and low-level reasoning
• A track record of owning a perception domain or subsystem end to end on a deployed system: you set the architecture, made the tradeoffs, and lived with the consequences through iteration in the field
• Production C++ for real-time inference on embedded or resource-constrained targets, plus Python for training pipelines and tooling
• Sound judgment on build-versus-buy, model-versus-classical, and where to spend compute, made under real latency and power budgets
• Engineering hygiene to move fast in a live codebase and the standards to hold a team to it:
  Git, testing discipline, CI/CD, clean abstractions
• Comfort debugging at the system level using logs, telemetry, and real-world data

• Experience with autonomous aircraft, UAS, or other aerial platforms
• Depth, stereo, or LiDAR fusion in production
• Real-time or embedded systems experience, including GPU or accelerator optimization
• Familiarity with ROS2 or similar robotics frameworks
• Experience mentoring engineers or leading perception work streams without formal management authority

Due to U.S. government contract requirements, this role is limited to U.S. citizens, U.S. permanent residents, or candidates from specific countries authorized under applicable export control regulations.

Skyways is an Equal Opportunity employer…