Computer Vision Software Engineer (Madrid)

Computer Vision Software Engineer (Madrid)

21 nov
|
Cactus
|
Madrid

21 nov

Cactus

Madrid

Cactus is a software development company, providing full-cycle services for the European market. More than 170 companies increased business productivity using solutions developed by Cactus’ team. The company has been regularly included in the listings of the best software development companies in Belgium, where we have our HQ, and Europe.About the roleCactus is hiring a Software Engineer with proficient Python and computer vision skills to build the 3 D and imaging pipelines behind our infrastructure-monitoring products, such as Tree Sight, our satellite and AI platform for vegetation-risk management along high-voltage power lines.You will turn satellite stereo imagery, Li DAR point clouds, RGB and thermal imagery into reliable, scalable insight for infrastructure operators. You will work with open-source 3 D tooling such as CARS, the CNES multi-view stereo pipeline that produces Digital Surface Models from satellite image pairs, and train the computer vision models that sit on top of that data.Beyond writing code, you will help shape how this data is used to monitor critical infrastructure: which products we build, what accuracy they need, and how they scale from a pilot area to a national network.What you'll do Build and operate photogrammetry and 3 D reconstruction pipelines (CARS or similar) that generate Digital Surface Models from satellite stereo imagery.Design and implement state-of-the-art multi-sensor data fusion algorithms combining Li DAR, satellite, RGB, multispectral and thermal data.Develop and improve our in-house point cloud classification algorithm, using KD-tree spatial search for geometric features and a Light GBM model for the final classification steps.Process, filter, classify and fuse large Li DAR point clouds with RGB, multispectral and thermal imagery, including co-registration and georeferencing.Design, train and evaluate computer vision models for detection (including YOLO-family models),



segmentation and change detection on aerial, satellite and thermal data, including hotspot and anomaly detection on infrastructure assets.Make pipelines fast and scalable: profiling, parallel and distributed processing, tiling strategies, and GPU acceleration where it pays off.Validate results against ground truth, such as Li DAR drone surveys, and define the accuracy metrics our customers rely on.Package work as maintainable, tested Python services ready for production and cloud deployment.Work with product and domain experts to decide how data can best improve infrastructure monitoring, and turn those ideas into prototypes. What we're looking for Proficient Python: clean, tested, performant code, with solid use of Num Py and the scientific Python stack.Proven computer vision experience, from classical methods (stereo, feature matching, geometry) to deep learning with Py Torch and object detectors such as YOLO.Hands‑on work with Li DAR point clouds (laspy, lazrs, Num Py, Sci Py KD‑trees) and with RGB, multispectral or thermal infrared imagery.Experience training, evaluating and improving CV and machine learning models (YOLO, gradient boosting such as Light GBM) on real‑world, imperfect data.A performance mindset: profiling, memory‑aware processing of large datasets, and parallel or distributed computing (Dask, multiprocessing, Ray or similar).Working knowledge of geospatial data: coordinate systems and transformations (pyproj), 2 D geometry (shapely) and rasters (rasterio).Daily use of AI‑assisted development, such as Claude Code, to write,



test, refactor and document code faster, while keeping full ownership of quality and correctness.A degree in computer science, telecommunications, geomatics, physics or a related field, or equivalent experience.Good written and spoken English; Spanish is a plus. Niceto have Direct experience with CARS, ASP (Ames Stereo Pipeline), Mic Mac or other multi‑view stereo tools.Earth observation background: Sentinel, Pléiades or other optical satellite data, and sensor models such as RPC.Radiometric thermal imaging: sensor calibration, emissivity and atmospheric correction, thermal orthomosaics, and hotspot analysis for electrical components or insulation.Point cloud deep learning (Point Net++, KPConv, sparse convolutions) or 3 D semantic segmentation.Work on infrastructure monitoring: power lines, vegetation encroachment, rail, roads or utilities.Cloud and HPC deployment: Docker, Kubernetes, AWS/Azure/GCP, or SLURM clusters.C++ or CUDA for performance‑critical code. You will join an AI-first engineering company working on funded space and Earth‑observation R&D, with real impact on how critical infrastructure is monitored. Ownership of core technical components from day one, and a say in the product direction.Work with satellite, Li DAR and drone data in projects with energy operators and European partners.Location: Seville, Spain (2–3 days a week in the office, the rest remote).Contract, salary range and start date: to be confirmed. What We Offer We value spontaneous interaction, co-creation, and co-ideation. That’s why we believe in hybrid work based at our office in Sevilla.A collaborative environment focused on ownership, growth, and impact.An environment to keep pace with AI-assisted tools.An open, international team of experts.Health insurance.Work‑related courses, certifications, and books. #J-18808-Ljbffr

📌 Computer Vision Software Engineer (Madrid)
🏢 Cactus
📍 Madrid

Postulate a este anuncio

Muestra tus habilidades a la empresa, rellenar el formulario y deja un toque personal en la carta, ayudará el reclutador en la elección del candidato.

Suscribete a esta alerta:

Recibe por email las nuevas ofertas de trabajo para: computer vision software engineer (madrid) / madrid

Suscribete a esta alerta:

Recibe por email las nuevas ofertas de trabajo para: computer vision software engineer (madrid) / madrid