A rigorous audit for an AI/ML paper, technical report, or research release. I identify where claims outrun evidence, where experimental details block reproduction, and where technical writing creates unnecessary reviewer friction. You receive: • Claim-to-evidence review of results, tables, and figures • Reproducibility checklist covering methods, configs, data, and code • Prioritized issue report with concrete fixes • Clarity edits for the abstract, contributions, and key technical sections • One follow-up revision round This service does not fabricate results, citations, or experiments. Send the manuscript plus any code, appendix, or result files you want included in scope.