
Technology
|
Defense Systems
|
YC W26
|
Valuation:
Undisclosed

Last Updated:
March 24, 2026

Builds modular strike and reconnaissance drones (Bandit, Banshee) paired with an AI-driven swarm control platform (Aerie) that enables single operators to coordinate autonomous drone swarms for military missions.
Seeing Systems has publicly showcased Bandit (budget-friendly FPV drone for training and one-way strike missions), Banshee (battle-proven FPV with fiber-optic guidance for EW-heavy theaters ensuring jamming immunity), and Aerie (intelligent swarm software for multi-platform operations). They have confirmed field testing with UK Royal Marine Commandos, Estonian Military, and European Armed Forces. Modular hardware with swappable payloads, sensors, compute, and comms supports strike, ISR, or swarm missions. Forward Deployed Engineers give each unit a point of contact, integrating feedback in hours.
Non-traditional signals suggest a strong push toward deeper agentic autonomy — moving beyond swarm coordination toward fully autonomous mission planning and dynamic threat response. The founders' backgrounds (Jane Street quant rigor + autonomous semi-truck field engineering) hint at sensor fusion and edge ML inference on-drone. GitHub and conference silence is consistent with classified or export-controlled work. Launching out of YC with customer traction, a product, and manufacturing capacity. Building presence in additional countries to service their needs.
Agentic Swarm Coordination: AI-driven multi-drone autonomous mission execution enabling a single operator to control an entire drone swarm in contested environments.
Instead of needing a pilot for every drone, one person tells the AI what the mission is and the swarm figures out how to do it together.
It's like being a restaurant manager who just says "busy Friday night, 200 covers" and the entire kitchen staff self-organizes who's on grill, who's on desserts, and who covers when someone burns their hand.
Battlefield Adaptive Autonomy: Real-time ML-driven flight path optimization and threat avoidance using live combat data from Ukraine deployments.
The drone teaches itself to dodge electronic jamming and enemy fire by learning from what happened to drones before it on the same battlefield.
It's like a delivery driver who doesn't just use Google Maps but actually remembers every pothole, speed trap, and road closure from thousands of previous drivers' dashcam footage—and reroutes before you even hit the problem.
Intelligent Target Recognition and Strike Optimization: ML-powered onboard computer vision for autonomous target identification, classification, and optimal strike angle computation on the Bandit drone platform.
The drone's onboard AI recognizes what it's looking at—vehicle, structure, decoy—and calculates the best angle of attack all by itself, even if it loses contact with the operator.
It's like a smart bowling ball that not only rolls itself down the lane but picks which pin to hit and adjusts its spin mid-roll for a guaranteed strike—even after you've already let go.
Matthew Le Maitre is a former Jane Street software engineer and top-ranked CS graduate from the University of Cambridge, where he worked at the Mobile Robotics Lab on autonomous systems and context-aware drones. Alex Le Maitre leads hardware and brings hands-on drone manufacturing and field engineering experience. The brothers combine elite quantitative software engineering with military-grade hardware expertise, with live battlefield feedback from Ukraine and NATO forces that no competitor can easily replicate. Close collaboration with UK Royal Marine Commandos led directly to the development of their systems.