International Journal of Scientific Research and Engineering Development

The volume of published research now outpaces any individual researcher's ability to read, compare, and synthesize it, making literature review one of the slowest and most error-prone stages of the research process. Single-agent large language model (LLM) assistants that attempt to search, read, summarize, and cite papers inside one undifferentiated reasoning pass are prone to fabricated citations, inconsistent coverage across repeated runs, and an inability to explain why a given paper was included or excluded. This paper proposes AutoResearch, a multi-agent architecture that decomposes automated literature review across five cooperating agents: an Orchestrator, three deterministic specialist agents (literature retrieval, paper parsing and extraction, citation mapping), and an LLM Synthesis Agent supported by a three-tier memory system. Agents communicate through the Model Context Protocol (MCP), a standardized client-server tool-calling layer, rather than bespoke point-to-point integrations with each academic data source. We present the complete system architecture, agent interaction sequence, communication protocol, memory architecture, deployment architecture, coordination algorithms, and a tagged case-study protocol for reporting agent behaviour consistently across evaluations. A worked case study demonstrates the pipeline end-to-end on a sample research query. Full benchmark-scale quantitative evaluation is scoped as an immediate next phase and is described here as a concrete, reproducible protocol rather than presented as completed results.

Jahnavi Somaraju, Meenakshi A, Nandini K, "AutoResearch: A Multi-Agent AI System for Automated Literature Review, Paper Summarization, and Citation Mapping" International Journal of Scientific Research and Engineering Development, V9(4): Page(506-519) May-June 2026. ISSN: 2581-7175. www.ijsred.com. Published by Scientific and Academic Research Publishing.