Nexus Research

Nexus Research is an AI-powered academic research platform that automates the entire literature review process from start to finish. Researchers, students, and curious minds can submit any research question and receive a comprehensive, structured analysis in minutes something that would traditionally take hours of manual reading, note-taking, and writing. At its core, Nexus Research runs a sequential multi-agent AI pipeline powered by Groq's ultra-fast LLaMA 3.1 models. The first agent, the Literature Extractor, queries the ArXiv API to retrieve up to ten relevant academic papers and enriches each one with a relevance score, a key contribution summary, and topic tags. The second agent, the Synthesis Engine, transforms those papers into a structured narrative synthesis complete with key findings, methodology overview, and limitations while simultaneously generating an interactive knowledge graph and a paper comparison matrix in parallel. The third agent, the Critique Agent, acts as a peer reviewer, surfacing research gaps, methodological biases, and concrete recommendations for future work. Results are streamed back to the browser in real time using Server-Sent Events, so users can watch each agent complete its task live. The frontend presents everything across four distinct interactive views: a Synthesis Tab with organized, color-coded research cards; a Mindmap Tab with a custom SVG radial knowledge graph that users can drag and pan; a Matrix Tab with a sortable heatmap table that scores each paper across multiple comparison dimensions; and a Chat Tab where users can ask follow-up questions to an AI assistant that is fully grounded in their specific research context. Built with React 18, TypeScript, Vite, TailwindCSS 4, and Supabase Edge Functions running on Deno, Nexus Research is entirely serverless, persistent, and scalable with all session data, paper sources, synthesis outputs, and agent logs stored in a Supabase PostgreSQL database for retrieval and reuse.