ThesisForge

ThesisForge helps students and early-stage researchers review their thesis work before submission. Instead of generating a thesis for the user, it focuses on quality control: finding unclear research gaps, weak citation support, methodology mismatches, under-explained results, and likely defense risks. Users create a thesis project, paste text or upload research materials such as PDF, DOCX, TXT, BibTeX, and CSV files, then run an asynchronous multi-agent review workflow. Specialized agents check literature coverage, research gap quality, citation alignment, methodology consistency, results interpretation, and defense readiness. The system shows visible agent progress, collaboration handoffs, and a final structured report with scores, priority fixes, action tasks, citation concerns, and likely panel questions. For judges, ThesisForge includes a safe synthetic demo project so the full workflow can be tested quickly without private thesis data. The project is built as a SaaS-style research review workspace using a Next.js frontend, FastAPI backend, Supabase authentication, Redis/RQ background jobs, Postgres, OpenAI-compatible LLM orchestration, and Band for agent collaboration and traceability.