Clevia

As LLM usage for assisted writing grows, hallucination of these models are creating a creating a corpus of drafts with incorrect or false citations that cannot be traced back to real sources. False or incorrect citations hurt knowledge credibility and create reputational risk for institutions and authors. If a paper includes a wrong or non-existent reference, downstream work can inherit it (literature reviews, meta-analyses, grant writing), and authors who cite contaminated references can put their own work and reputation at risk. In the worst cases, that can lead to corrections or retractions. I built Clevia to make citation verification a fundamental step before you submit or share. It extracts citations from your papers or drafts, resolves them to real records (DOIs + authors/year/venue), and flags missing or inconsistent references. With direct links back to sources and a clean, export-ready bibliography in one workspace, you can write and cite with confidence, eliminating retraction and reputation risk. I’d love feedback on: - The most common failure mode you see (wrong metadata, wrong source, missing DOI, fabricated refs) - Where verification should run in your workflow (on import, during writing, or before export) - How Clevia helps you research and write faster and better?