Experiences
Led the deployment of scalable cloud services powering AI applications, exposed via REST APIs. Designed and built serverless, cloud-native architectures to support internal ML tools and client-facing solutions.
Designed distributed architectures for rule-based systems and high-volume ETL pipelines in anti-fraud and AML domains. Developed internal libraries and NLP-based automation services, deployed across multiple environments as REST APIs.
Researched and trained deep learning models for object detection and instance segmentation in insurance claim processing. Contributed to the system architecture enabling production deployment and cost estimation based on image analysis.
Built ETL pipelines and analytical tools on distributed systems using Scala, Python, and Spark. Focused on data extraction, transformation, and analysis to support large-scale financial models and internal analytics platforms.
Education
Developed an industrial thesis applying Deep Learning and Computer Vision to assess car damage from images. Combined object detection and instance segmentation models to identify vehicles, detect components, and assess damage severity. Integrated business rules to estimate repair costs from model outputs.
Explored semantic representation of web pages by combining graph-based random walk algorithms with textual analysis of HTML content. Learned vector embeddings of pages and evaluated the approach through unsupervised clustering with promising results.