Pau is currently unavailable
Pau Carré Cardona
Verified Expert in Engineering
Machine Learning Developer
Palma de Mallorca, Spain
Toptal member since July 4, 2023
Bio
Pau is a seasoned data scientist with remarkable computer vision and NLP expertise and a proven track record of developing large-scale AI systems and implementing them at scale. As an invited speaker at the prestigious O'Reilly AI Conference in New York, he showcased his ability to captivate and inform technical and business audiences. Pau excels in communication, fostering effective collaboration across diverse disciplines and cultural backgrounds, and delivering high-value software solutions.
Portfolio
Zendesk
PyTorch, Machine Learning, Data Science, Python, Artificial Intelligence (AI)...
Coalescent Mobile Robotics
Robot Operating System (ROS), Python, Artificial Intelligence (AI), Frameworks...
Zalando
Deep Learning, Machine Learning, Python, Artificial Intelligence (AI)...
Experience
- Linux - 15 years
- Python - 10 years
- PyTorch - 6 years
- Machine Learning - 6 years
- Deep Learning - 6 years
- Data Science - 5 years
- Robotics - 2 years
- Robot Operating System (ROS) - 2 years
Preferred Environment
Linux
The most amazing...
...eCommerce image similarity feature I've worked on pioneered the adoption of AI in the fashion industry, starting as early as 2016.
Work Experience
Senior Machine Learning Engineer
2021 - PRESENT
Zendesk
- Automated ticket reply templates to improve the productivity of agents.
- Performed machine learning improvements in search to improve semantic search.
- Developed machine learning services, data science, and deep learning code.
Technologies: PyTorch, Machine Learning, Data Science, Python, Artificial Intelligence (AI), Frameworks, Computer Vision
Robotics Engineer
2019 - 2020
Coalescent Mobile Robotics
- Implemented real-time human detection and avoidance with exceptional precision utilizing advanced Intel D435 sensors, yolact++, PCL, and OpenVDB and STVL technologies.
- Achieved reliable robot localization through the seamless fusion of scan data and advanced object detection algorithms using the power of ROS2.
- Revolutionized automatic robot localization by harnessing the potential of high-resolution cameras. Employed the groundbreaking Facebook AI Similarity Search (FAISS) in the embedding space.
- Contributed to navigation by leveraging the robust capabilities of Nav2, including odometry and controller modules, to achieve unparalleled accuracy in robot navigation while implementing advanced behavioral planning using behavior trees.
Technologies: Robot Operating System (ROS), Python, Artificial Intelligence (AI), Frameworks, Computer Vision, Image Processing
Applied Data Scientist
2018 - 2020
Zalando
- Developed style-similarity to complete fashion collections using style-transfer style embeddings. Automated discovery and classification of fashion trends using clustering on hierarchical embedding spaces.
- Implemented few-shot classification using variational autoencoders and GANs.
- Built a groundbreaking framework that seamlessly combines recommendations from diverse sources and distributions into a unified, homogeneous system.
Technologies: Deep Learning, Machine Learning, Python, Artificial Intelligence (AI), Frameworks, Computer Vision, Image Processing
Applied Data Scientist
2015 - 2018
Hudson's Bay Company
- Created an unsupervised image similarity system to generate automated product recommendations.
- Built systems to predict product facets using deep neural networks based on NLP and image detection.
- Expanded pixel-level image segmentation for garments.
- Developed a rule-based recommendation system as part of my role.
Technologies: Deep Learning, Machine Learning, Data Science, Python, Artificial Intelligence (AI), Frameworks, Computer Vision, Image Processing
Senior Software Engineer
2015 - 2015
Zynga
- Developed .NET infrastructure and tooling to test video games for mobile devices automatically.
- Built a Java server to orchestrate the execution of integration tests in mobile devices.
- Maintained and monitored the testing infrastructure.
Technologies: .NET
Team Lead Software Engineer
2012 - 2014
Nomad Automate
- Implemented and spearheaded the core development of Nomad OSS, a specialized geographic information system (GIS) for microwave networks tailored specifically for telecom operators, leveraging Java.
- Architected and led the core development of DCN Wizard, an advanced Java desktop application to configure and validate data communication networks (DCN) through an intuitive visual representation of the network.
- Pioneered the creation of a cutting-edge, model-driven Java (GWT) web front end for Nomad OSS to enable telecom operators to manage and optimize their network infrastructure seamlessly.
Technologies: Java
Head of IT
2009 - 2011
Municipality of Pollença
- Managed networking and system administration employees.
- Developed several internal Java web applications to automate internal government administrative procedures.
- Automated and self-verified Windows Server and network administration processes achieved through cutting-edge mechanisms.
Technologies: Microsoft Servers, Java
Security Software Engineer
2007 - 2009
Soffid
- Implemented model-driven development techniques to create a highly scalable Java EE web application capable of efficiently managing identity and access control for corporate systems.
- Engineered distributed Java agents with robust synchronization capabilities, enabling seamless propagation of credentials and access control rules across various interconnected systems.
- Built a Java desktop application that leverages smart card technology to securely sign content and validate digital signatures, ensuring the utmost integrity and authenticity of data.
Technologies: Java, Cryptographic Protocols, Accredible Credential API
Experience
AI for Fashion Industry
This project seamlessly integrates machine learning and deep learning technologies within the fashion and eCommerce sectors. The primary objectives encompass automated tagging and image similarity-based recommendations. Remarkably, this initiative pioneered the adoption of artificial intelligence in the industry, starting as early as 2016.
Robot Arm From Scratch
https://www.youtube.com/watch?v=uqI_j1Ziaoc&list=PLZx8jDELltyFXH5rgVqxvLg4pokd7oDY3&index=2I designed and constructed a SCARA and a mobile robot, encompassing a wide range of tasks such as mechanical design, 3D printing, electronics, assembly, embedded programming, artificial intelligence, kinematics, and control.
The mechanical design was executed proficiently using Fusion 360, while computer vision and deep learning techniques were applied to understand 3D space comprehensively. I also employed multicore FreeRTOS and ESP32 for the embedded programming aspect of the project.
The mechanical design was executed proficiently using Fusion 360, while computer vision and deep learning techniques were applied to understand 3D space comprehensively. I also employed multicore FreeRTOS and ESP32 for the embedded programming aspect of the project.
Education
2001 - 2006
Master's Degree in Computer Science
Universitat Politècnica de Catalunya - Barcelona
Certifications
FEBRUARY 2019 - PRESENT
Modern Robotics, Course 5: Robot Manipulation and Wheeled Mobile Robots
Northwestern University (via Coursera)
FEBRUARY 2019 - PRESENT
Modern Robotics, Course 2: Robot Kinematics
Northwestern University (via Coursera)
FEBRUARY 2019 - PRESENT
Modern Robotics, Course 1: Foundations of Robot Motion
Northwestern University (via Coursera)
OCTOBER 2016 - PRESENT
Robotics: Perception
University of Pennsylvania (via Coursera)
Skills
Libraries/APIs
PyTorch, Accredible Credential API
Tools
Autodesk Fusion 360
Languages
Python, Java
Platforms
Linux, Amazon Web Services (AWS)
Frameworks
.NET
Other
Machine Learning, Data Science, 3D Printing, Deep Learning, Artificial Intelligence (AI), Frameworks, Computer Vision, Image Processing, Robot Operating System (ROS), Robotics, Software Engineering, Mathematics, Physics, Microsoft Servers, Cryptographic Protocols
Collaboration That Works
How to Work with Toptal
Toptal matches you directly with global industry experts from our network in hours—not weeks or months.
1
Share your needs
Discuss your requirements and refine your scope in a call with a Toptal domain expert.
2
Choose your talent
Get a short list of expertly matched talent within 24 hours to review, interview, and choose from.
3
Start your risk-free talent trial
Work with your chosen talent on a trial basis for up to two weeks. Pay only if you decide to hire them.
Top talent is in high demand.
Start hiring