Reproduce “ORAM.AI: A Multi-Agent Chatbot With RAG-Based Reasoning and MCP Integration for Adaptive Fitness Recommendation” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Konstantinos I. Roumeliotis, Orestis N. Zestas, Kyriakos D. Kyriakou, Dimitrios N. Soumis, Sofia H. Kapellaki, Kyriaki Seklou, Nikolaos D. Tselikas
ORAM.AI routes fitness queries through a multi-agent system that combines conversation context, retrieval of exercise records, and a fine-tuned advisor. In a 15-query ablation, MCP retrieval alone produced nonzero verbatim exercise-name hits, while fine-tuning alone had the highest safety/form cue count. The combined system was slower than a single-LLM baseline; these text-based measures do not establish safety.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.