Who we are and what you'll build

Autonomous drones need two things to work: a way to see and a way to decide where to go. Thalanor's R&D; team builds both. If turning advanced mathematics and deep learning into presente flight paths for Unmanned Aerial Systems (UAS) sounds like your kind of challenge, read on. This isn't model tuning. You'll take part in designing the "brains and eyes" of autonomous drones from first principles, covering guidance, interception and computer vision.

The position

Thalanor is a small, highly technical R&D; team, and you'd be part of it. The work sits where Optimal Control Theory, computer vision and autonomous navigation for UAS come together. The aim is to push guidance and vision algorithms beyond the state of the art, especially for interception and for detecting objects in complex environments.

Vision: you'll apply deep learning–based object identification, including Long‑Short‑Term Memory vision models wherever they add value, to make detection more accurate and more robust.

Autonomy: you'll help develop guidance algorithms rooted in Optimal Control Theory, defining interception strategies and refining trajectories for real drones flying in the field.

Day to day

- Start from the latest literature, then adapt, combine or invent computer vision and guidance methods that actually work in practice.
- Write the code yourself, moving fast from experiments to working modules that plug into our drone systems.
- Work with interception trajectories and statistical filtering (for example Kalman Filter, Extended KF and Unscented KF) to make the whole system more stable and better performing.
- Team up with other engineers in a startup‑like setting, where autonomy, proactivity and the ability to "figure things out" are essential.





Expect to spend most of your time on hands‑on R&D;: reading papers, testing ideas and turning the best ones into robust, maintainable code. If you like switching between theory and implementation, and you want to see your work fly – literally – this role fits.

Your profile

- A strong quantitative background (mathematics, physics, engineering or similar), with genuine enjoyment of mathematical thinking.
- Solid programming skills, ideally in languages and frameworks commonly used for numerical algorithms and deep learning.
- Experience in computer vision and/or deep learning for object detection or related tasks; you know how to choose, train and evaluate models, and how to debug them when reality doesn't match the theory.
- Familiarity with at least one of the following areas, or the ability to ramp up quickly: Optimal Control Theory, Long‑Short‑Term Memory vision models, Kalman‑family filters.
- A proactive, startup‑oriented mindset: you like ownership, can manage your own workload, and you're happy to explore problems that don't have a predefined solution.

We don't need you to arrive with deep expertise in every algorithm we use. What matters is your ability to understand and extend complex methods, plus the drive to build something new rather than simply apply off‑the‑shelf solutions.

What makes this role worth considering

This is frontier technology in autonomous flight, at a company where R&D; is central to the product, not an afterthought. Thalanor develops advanced defence systems built for protection and deterrence, not offensive weapons, and your work will run in live systems, not just simulations.

Next step

Apply through the link. If you'd like to talk about any aspect of the role first, reach out to our Recruiting Manager.

📌 R&D Engineer - Computer Vision and Guidance Systems (Madrid)
🏢 Talento Cientifico
📍 Madrid