Automated fraud detection
System for organizing & training models and making predictions. Consisting of 3 components (REST API, Dataset service, Machine learn...
System for organizing & training models and making predictions. Consisting of 3 components (REST API, Dataset service, Machine learning service). Designed to the CQRS model where the REST API converts incoming HTTP requests into events. The main focus points were extensibility and data consistency; therefore, key strengths in data classifications (via the attrs library) and model abstractions (sqlalchemy and attrs). Machine learning component works with SKLearn pipelines (co- developed with TU Delft data scientists) and Pandas data sets.
PostgreSQL
Python 3
Microservices
View more
PostgreSQL
Python 3
Microservices
RESTful API
View more
Digital templating service
Web application designed and developed for reading design files (sketch, indesign), offering an in-browser editor of these files, ex...
Web application designed and developed for reading design files (sketch, indesign), offering an in-browser editor of these files, exporting to digital formats (JPG, PDF, GIF) and publishing to marketing channels (facebook, twitter, instagram). Interface runs via GraphQL to Python Lambda functions via AWS API Gateway and storage in DynamoDB. Among others, users were TUI, Capital One, Volvo and Amadeus.
Python 3
React
GraphQL
View more
Python 3
React
GraphQL
AWS DynamoDB
AWS Lambda
View more