Vasilis Liapas

-Modernized legacy banking APIs (WSDL, XSD, XML, C#, Java) by converting them to OpenAPI. Identified system dependencies, enhanced documentation and applied AI-based similarity to highlight redundancies, enabling data-driven decommissioning decisions. -Acted as a technical bridge between the AI and Digital Engineering teams, guiding on the architectural decisions on API transcoding logic, GenAI integration, and Python implementation strategy. -Developed a RAG application that managed contracts between businesses, enabling efficient retrieval of relevant information and supporting advanced search and analysis capabilities. -Modified, obfuscated and deployed a Python tool for a major insurance company, making it compatible with their systems and databases. Set up scheduled scripts to run automatically in their secure environment. -Automated the testing of small LLMs on selected datasets and conducted analysis and comparison of experimental results with quantized Large Language Models. -Developed a synthetic time series dataset for the experimentation of pretrained and traditional machine learning models. Developed the visualization on the time series forecasting experiment results. Supported in the implementation of a Risk Analysis Pipeline. -Contributed on the development of an environment for data lineage analysis in IICS data flows.