• Working on computer vision algorithms for the insurance industry, mainly deep learning based classification, regression and detection models. Performing applied research work to allow insurers to automate the process of claim settlement and fraud detection.• Research challenges include training multi-instance and multi-task models in an environment of weak supervision with noisy multimodal data.• Close work with the product team and with clients, implementing an efficient production monitoring platform and feedback based model improvement loops.• Publication: “AutoFraudNet: A Multimodal Network to Detect Fraud in the Auto Insurance Industry” - The AAAI-2023 Workshop On Multimodal AI For Financial Forecasting.• Technological stack: Python, TensorFlow, Keras, Pandas, Numpy, Sklearn, OpenCV, AWS, GCP etc.

• Developed state-of-the-art Machine Learning algorithms for bespoke computer vision tasks applied to the retail industry such as object detection, tracking and classification.• Research in an environment of several challenges such as scarce and imbalanced data, low resolution images, low frame-rate videos, and a strict computational budget.• R&D on AWS and GCP platforms, mostly using Python, Keras and TensorFlow.

Research in the field of Deep Learning under the supervision of Professor Miguel Rodrigues with focus on information theoretic approaches for the analysis of deep neural networks, specifically in the context of semi-supervised transfer learning."Lautum Regularization for Semi-supervised Transfer Learning" - Workshop on Statistical Deep Learning in Computer Vision, ICCV 2019.https://arxiv.org/pdf/1904.01670

Algorithms engineer in the WiFi department:• Developed signal processing and machine learning algorithms and calibrations for performance optimization. The role included end-to-end development of WiFi physical layer communication algorithms from Matlab design to system integration and testing.• Developed system tests and implemented algorithms in C#.

• Algorithms & system engineer in the space division, in charge of the verification and validation of a newly developed satellite system.• Worked on the design of space mission control algorithms.