ResearchPilot: A Local-First Multi-Agent System for Literature Synthesis and Related Work Drafting

TL;DR

ResearchPilot is an open-source multi-agent system that assists with literature reviews by retrieving, extracting, synthesizing, and drafting related work sections from scientific papers in a local-first architecture.

Contribution

It introduces a system design for a transparent, local-first research assistant that integrates multiple tools for literature synthesis and drafting, emphasizing engineering tradeoffs over algorithmic novelty.

Findings

Successfully retrieves papers from Semantic Scholar and arXiv

Synthesizes cross-paper patterns and drafts related work sections

Operates effectively in a local-first, self-hosted environment

Abstract

ResearchPilot is an open-source, self-hostable multi-agent system for literature-review assistance. Given a natural-language research question, it retrieves papers from Semantic Scholar and arXiv, extracts structured findings from paper abstracts, synthesizes cross-paper patterns, and drafts a citation-aware related-work section. The system combines FastAPI, Next.js, DSPy, SQLite, and Qdrant in a local-first architecture that supports bring-your-own-key model access and remote-or-local embeddings. This paper describes the system design, typed agent interfaces, persistence and history-search mechanisms, and the engineering tradeoffs involved in building a transparent research assistant. Rather than claiming algorithmic novelty, we present ResearchPilot as a systems contribution and evaluate it through automated tests and end-to-end local runs. We discuss limitations including external API rate limits, abstract-only extraction, incomplete corpus coverage, and the lack of citation verification.