
ThesisForge
Multi-agent thesis quality control that finds what's weak before submission
Problem
Thesis work is scattered across disconnected tools (Scholar, Zotero, Word/LaTeX, spreadsheets, supervisor emails), so students often discover a weak research gap, unsupported claims, or a mismatched methodology only after the thesis is already assembled.
Solution
Built a SaaS-style review workspace where users upload thesis text/PDF/DOCX/BibTeX/CSV and a visible 7-agent pipeline (literature coverage, research gap, citation alignment, methodology consistency, results interpretation, defense prep, report generation) runs asynchronously via Redis/RQ, using Band as the agent collaboration layer, and returns a scored report with priority fixes and defense questions.
- Stack
- Next.jsTypeScriptFastAPIPythonPostgreSQLRedisRQSupabase AuthOpenAIBand
- Last updated
- September 2026
- GitHub
- 0 stars · last commit September 2026