Ralph ToumaHome Work
Work
Projects and roles, written out properly
Four pieces of work: two independent machine learning projects, and two roles where the work shipped to someone. Each one explains the problem, the approach and what was actually measured.
Case studies
Personalized privacy-aware federated weather forecasting
I built a federated learning system that forecasts next-day weather across city-level clients, trained on non-IID station data in Flower's simulation framework, and benchmarked it against centralized, local-only and FedAvg baselines.
Enterprise RAG assistant for technical documents
I built a retrieval-augmented generation assistant over technical PDFs and internal documents: ingestion and chunking, embedding generation, vector storage for semantic retrieval, and citation-grounded answers.
Machine learning research at Green-Communications
A research internship in Paris (July to November 2025): federated and clustered federated learning applied to weather forecasting and heterogeneous distributed systems.
Commerce engineering and analytics at BYND Network
Shopify storefronts and custom apps for 15+ clients, performance dashboards for internal teams, and data analysis that improved storefront conversion by 30%.