AITAN is building the operating platform for multi-drone warfare, a battle-proven system that augments, deploys, and orchestrates autonomous drone fleets across any mission, any environment, and any vendor. With over 500,000 hours of frontline operational use and TRL-9 validation, our software is not a prototype. It runs in combat, every day, shaping outcomes in some of the most contested and complex environments on earth.
We unify drone control, edge-AI processing, and mission execution into a single operational loop, from detection to decision to strike. Our platform gives a single operator the ability to command and coordinate an entire fleet in real time: persistent aerial surveillance, automated target recognition, shared tactical picture, and seamless handoff from ISR to kinetic response, without manual delay.
We are expanding our US engineering team in the New York metropolitan area. Our office is easily accessible from New York City. If you want to work on technology that matters, not in a lab, but validated in the field, this is the role for you.
What you'll do
We are looking for an exceptional Computer Vision Engineer to join our research team and own the perception layer that powers our autonomous aerial systems. You will design and build the systems that give our drones and the operators behind them a real-time, shared understanding of the operational environment. Your algorithms will run at the edge, in contested airspace, with real operational stakes.
You will take perception capabilities from concept and simulation through live flight experiments to production deployment, iterating rapidly on real-world flight data and operational edge cases. The quality of your work directly determines what warfighters can see, understand, and act on.
Design, implement, and continuously improve vision-based situational awareness systems that give aircraft and ground operators a unified, real-time understanding of the operational environment
Build and maintain a shared world model across the fleet, fusing perception data from multiple aircraft into a coherent 3D representation that all agents can query and act on
Implement object detection, tracking, and classification pipelines covering dynamic obstacles, terrain, infrastructure, and other aircraft
Develop semantic scene understanding to power onboard autonomy and off-board mission planning
Design multi-agent perception architectures that aggregate observations across the fleet, resolve conflicting views, and maintain a consistent, up-to-date environmental state
Build perception that functions in GNSS-denied and EW-contested environments: your systems must work when GPS is unavailable and comms are degraded
Combine classical computer vision with modern deep learning to deliver robust, low-latency perception across diverse terrain, lighting, and atmospheric conditions
Own the full algorithm lifecycle: design → simulation → onboard and offboard deployment → production tuning and ongoing field improvement
Collaborate directly with flight controllers, system engineers, and operations teams, iterating rapidly on real-world flight data and operator feedback
Requirements
M.Sc. in Computer Science, Electrical Engineering, Robotics, Aerospace Engineering, or a closely related discipline
4+ years of experience developing computer vision, perception, or autonomy algorithms for production systems
Strong command of both classical computer vision and modern deep learning; you know when to use each
Hands-on experience with object detection, SLAM, visual odometry, or 3D scene understanding
Proven track record of taking algorithms from research into real-world environments under genuine operational constraints
Proficiency in Python and C++; experience deploying algorithms under real-time and latency constraints
Comfortable working closely with hardware, sensors, and embedded compute platforms
Experience with modern neural network architectures and deploying trained models in constrained environments
Strong fit if
You have experience with drones, aerial systems, or autonomous robotics, ideally in a defense or field-tested context
You've built perception systems for GNSS-denied or GPS-degraded navigation (visual-inertial odometry, optical flow)
You have experience with multi-sensor fusion, sensor calibration, or sensor-agnostic perception pipelines
You've participated in simulation-to-real workflows and live flight testing programs
You have experience deploying models on edge compute hardware such as NVIDIA Jetson, Qualcomm, or similar
You've worked in a fast-paced defense tech or deep-tech startup environment where field feedback drives rapid iteration