{"repo":"terence-lim/quant-agents","free":true,"listed":false,"github":"https://github.com/terence-lim/quant-agents","clone":"git clone https://github.com/terence-lim/quant-agents.git","description":"An Agentic AI Framework for Empirical Research in Quantitative Finance","language":"Python","stars":15,"topics":["agentic-finance","factor-research","mcp","pydantic-ai","quantitative-finance","streamlit"],"license":"MIT","category":"trading","readme_excerpt":"Quantitative Research Assistants with Financial Tools and Intelligence (QRAFTI) (c) Terence Lim 2026 A tool-augmented multi-agent framework designed to emulate a quantitative research team, QRAFTI integrates (1) an empirical research Python toolkit built for panel data, (2) Model Context Protocol (MCP) tool servers that expose data- and factor-manipulation operations as callable tools, (3) specialized Pydantic-AI agents for factor research, standardized reporting, and customized code writing and execution, and (4) Streamlit interface for tracing tool calls and visualizing data artifact connections. It demonstrates the use of LLM and agentic capabilities to simplify and strengthen quantitative research workflows. Lim, T., Muthuraman, K., & Sury, M. (2026). QRAFTI: An agentic framework for empirical research in quantitative finance [Preprint]. arXiv. https://arxiv.org/abs/2604.18500 Internet Appendix: documentation and results QRAFTI Usage Examples The following prompts illustrate how a user can work with QRAFTI for replication studies and autonomous factor research (see Internet Appendix B for full conversation histories and traces) 1) Replicate Fama-French HML-style workflow 2) Replicate JKP-style price momentum factor workflow 3) Autonomously generate and test a new factor idea Project Structure Python modules for running multi-agent quantitative research workflows, including data/factor services, user access interface, and empirical analysis tooling (see Internet Appendix A","default_branch":null,"files":null,"tree":[],"storefront":"/r/terence-lim","claimed":false,"request_supported":{"post":"https://gitbuyer.com/r/terence-lim/quant-agents/request-supported","requests":0},"note":"indexed from public GitHub; nothing is for sale on this page. Clone it from GitHub. Paid listings live at /search."}