Nermine Kabadou

• Built an LLM-based startup scoring system using unstructured data and investor feedback, replacing manual first-pass screening for 20 startups per month. • Streamlined the startup-investor matchmaking process, reducing manual evaluation time by 40%.

• Implemented an end-to-end fatigue prediction pipeline using time-series preprocessing and a softmax-based classification model to classify athlete fatigue levels from multi-sensor data.

• Implemented new website features and enhanced existing functionalities, which improved user experience. • Conducted thorough testing protocols that reduced web application errors by 30% and improved reliability and user engagement.

Created a modular, serverless infrastructure using Terraform and AWS services, to automate intern resource provisioning which eliminated 90% of repetitive manual steps for managers.