Deep Learning for Decision-Making & Image Processing

Ravenwits is a deep-tech spin-off from Universidad Politécnica de Madrid and Universidad Complutense de Madrid specialising in the application of advanced deep learning and stochastic optimisation to high-stakes, real-time decision-making problems. Our core R&D capabilities — convolutional neural networks for spatio-temporal data, gradient boosting ensembles, and reinforcement learning for sequential optimisation — translate directly into defence-relevant domains including autonomous systems, aerial image recognition, and sensor fusion.

Our team has hands-on experience developing and deploying AI models that process complex multimodal data streams and generate actionable outputs under uncertainty, currently validated in the electricity sector at national grid scale. These same architectural foundations underpin applications such as:

• Drone image recognition — object detection, classification, and change detection from UAV imagery using CNN-based pipelines

• Autonomous flight decision-making — real-time AI inference for path planning, obstacle avoidance, and mission-adaptive control

• ISR & situational awareness — pattern recognition and anomaly detection in aerial and remote sensing data