Ricardo is a Lead Software Engineer with 9 years of experience developing and deploying machine learning solutions using Python, TensorFlow and PyTorch. He has worked on a wide range of projects projects involving Deep Learning, Computer Vision, Natural Language Processing, Large Language Models, and Audio/Speech Processing — with applications ranging from human pose estimation and face recognition to acoustic source classification. Over the past year, he has also been working with AWS, leveraging cloud infrastructure to build and deploy scalable ML systems. Throughout his career, he has contributed to the development of intelligent systems in both research and industry contexts, building robust models and tools that support real-world impact.

He holds both a Bachelor's and a Master's degree in Computer Science from the Universidade Federal de Pernambuco (UFPE), with academic research focused on human pose tracking and dataset creation for pose estimation. During this time, he led multiple applied research initiatives, publishing peer-reviewed papers and collaborating with industry partners. His academic journey also includes work in Virtual and Augmented Reality, combining AI with immersive technologies to enhance user experience in healthcare and the arts.

Driven by curiosity and a passion for problem-solving, Ricardo is committed to building impactful AI systems that bridge the gap between research and real-world needs. He brings a hands-on, detail-oriented mindset to his work, always striving for clarity, robustness, and innovation. For him, technology is not just a tool — it's a way to empower people, improve lives, and explore new possibilities through intelligent solutions.

Technologies

• Python, Machine Learning, Computer Vision, LLMs, Deep Learning, TensorFlow, PyTorch, AWS, OpenCV, scikit-learn, TensorFlow Lite, Librosa, Git, Docker, Jira, MongoDB, Bash, SQL, CI/CD, Data Science, Transformers, Generative AI, NLP, Code Review, Scrum, Kanban

Languages

• Portuguese (native), English (professional proficiency)

Lead Computer Vision Engineer @ FCx Labs (Jan. 2025 - Current)

As a Lead Computer Vision Engineer at FCx Labs, I was involved in the development and enhancement of e-commerce product catalogs to improve customer experience through high-quality visual content. My work focused on designing and deploying computer vision solutions that automated and optimized the way product data was processed, displayed, and categorized across digital platforms.

• Technical leadership of a team of four Computer Vision developers, providing technical guidance and delegating tasks effectively to meet project goals.
• Collaboration with stakeholders and cross-functional business teams to define project scope, assess feasibility, and gather detailed technical requirements.
• Design and implementation of APIs and services for core computer vision tasks such as background removal, image captioning, product categorization from images, automatic image quality validation, virtual staging of products, and visual-based classification.
• Architecture of data pipelines for importing and processing products into the catalog, ensuring consistency and scalability across the system.
• Deployment of computer vision algorithms using AWS infrastructure, ensuring reliability, scalability, and performance in production environments.

• Standardized and automated the catalog image onboarding process using computer vision algorithms, reducing manual effort and errors, and bringing processing times down to just a few minutes.
• Deployed an image-based product categorization model capable of accurately classifying thousands of products with an accuracy of approximately 90%, enhancing the speed and precision of catalog organization.
• Built a cost-effective batch image inference pipeline on AWS SageMaker, processing large datasets at a cost on the order of a few cents to a few dollars per batch.

Skills: Python | AWS | Git | Docker | Rancher | TensorFlow | PyTorch | OpenCV | MongoDB | FastAPI | Jira

Applied Machine Learning Scientist @ SiDi (Jan. 2021 - Jan. 2025)

As an Applied Machine Learning Scientist at SiDi, I was part of the Speech Processing team, developing machine learning and deep learning solutions to bridge the gap between academic research and real-world applications. I focused on designing and deploying models for embedded systems, targeting Samsung’s mobile and IoT devices, with emphasis on efficient, production-ready solutions in audio, speech, and natural language processing.

• Engagement in regular meetings with international stakeholders to present experimental results, discuss technical challenges, and align development efforts with business goals.
• End-to-end development of machine learning models, including training, validation, testing, deployment, parameter tuning, and retraining, tailored to speech and audio applications.
• Design of lightweight and optimized models for embedded deployment, focusing on size, latency, and memory efficiency using techniques such as quantization, pruning, and TensorFlow Lite.
• Direction and execution of research initiatives resulting in scientific publications and patent filings in audio and speech processing.
• Delivery of internal technical talks to promote team-wide knowledge sharing and skill development.

• Built an embedded keyword spotting system with 95% accuracy, robust to noise and reverberation, deployed in the market and used by millions of users.
• Published three peer-reviewed scientific papers on acoustic source classification, introducing novel methodologies and benchmarks.
• Co-authored 1 INPI patent.
• Developed a customizable, fully embedded keyword detection system for user-defined wake words, achieving 92% accuracy and successful deployment on commercial devices.

Skills: Python | TensorFlow | Git | Docker | TensorFlow Lite | TFLite Micro | Jira

Academic Researcher @ Voxar Labs (Aug. 2016 - Aug. 2020)

As an Academic Researcher at Voxar Labs, I contributed to projects at the intersection of computer vision, machine learning, and virtual/augmented reality. My work combined foundational research with applied innovation, often in collaboration with industry partners to deliver research-driven solutions with real-world impact. I explored a wide range of topics, from human pose estimation to thermal-based bat tracking and immersive applications, leading to scientific publications.

• Academic research in computer vision, machine learning, and related fields, contributing to theoretical and practical advancements.
• Collaboration with industry to align research objectives with business needs and deliver research as a product.
• Investigation and evaluation of state-of-the-art computer vision techniques with a focus on feasibility and real-world applicability.

• Published 8 peer-reviewed papers across diverse fields such as human pose estimation and augmented reality.
• Built a flexible framework for rapid creation of human pose estimation datasets, significantly reducing data collection and annotation time.
• Developed a virtual reality application designed to guide ballet dancers through the learning and practice of fundamental arm positions.
• Led research on 3D object reconstruction from RGB images.
• Developed and deployed a tool for visualizing bat tracking data from thermal videos, enabling new insights in ecological research.
• Researched robust face recognition methods for varied image conditions.
• Designed and validated a proof-of-concept augmented reality system to support physiotherapy rehabilitation, demonstrating its potential for patient engagement and remote therapy.
• Successfully transferred research outcomes into commercial contexts through partnerships with private companies.

Skills: Python | Keras | PyTorch | OpenCV | C++ | Unity | Git | Docker | LaTeX

M.Sc. in Computer Science (Aug. 2018 - Jul. 2020)

• Informatics Center (CIn), Federal University of Pernambuco (UFPE)
• Master Thesis: HuTrain: um Framework para a Rápida Concepção de Datasets de Estimação de Poses Humanas

B.Sc. in Computer Science (Apr. 2014 - Jul. 2018)

• Informatics Center (CIn), Federal University of Pernambuco (UFPE)
• Undergraduate Thesis: Rastreamento de Poses Humanas a Partir de Entradas RGB