Personal details
Éderson S.
Summary
Python | Scala | LLM | Langchain | Machine Learning | Data Scientist | Computer Vision
Passionate Data Scientist and Computer Vision Engineer with a proven track record of leveraging machine learning to create value-driven solutions that enhance company outcomes and customer experiences. Specializing in insurance underwriting models, I have demonstrated adaptability and a keen interest in exploring new domains.
My expertise lies in building robust Python and Scala applications for diverse purposes, including production line inspections where I develop applications that capture real-time images from cameras, extracting crucial data for quality assurance and production metrics.
Key Areas of Expertise:
Data Science: Proficient in both Supervised and Unsupervised Learning, with a strong background in recommenders, forecasting, classification, and anomaly detection.
Computer Vision: Skilled in image classification and object detection using Pytorch and TensorFlow, and experienced in utilizing OpenCV for various vision tasks.
LLM: Utilizing LLM in conjunction with Langchain for RAG systems, focusing on task automation and interactions with APIs.
Knowledge in: MLOps, Deep Learning, NLP, BI with Power-BI, and Insurance Underwriting Models.
Main Tools: Python, Scala, Jupyter Notebooks, VS Code, Git, Numpy, Pandas, Sklearn, Pytorch, Airflow, DVC, MLFlow, Docker, OpenCV.
Work Experience
Working at NuSeguros, I develop Data Science solutions, which include training underwriting machine learning models, A/B tests and many data analysis to improve our internal services and products, making them even better and suitable for our customers. I also present all results and projects to stakeholder, including General Manager, external partners and all our business unit. My main tools are: Mainly we use platforms Databricks and Kubeflow. Programing languages: Spark Scala and Python.
Using the state of the art algorithm of detection, classification and segmentation in combination with image processing we built Computer Vision food inspection systems capable of checking thousands of products per hour. Main tools and frameworks: Yolov5, Pytorch, Tensorflow, Git, Kubernetes, Detectron2, OpenCV, Python, DVC, Airflow