Guoleme: Workplace AI Coach

Problem: high-stakes workplace reviews

The hard part is not writing a plan; it is defending value and handling pushback in interviews, promotions, salary talks, reviews, roadshows, and academic defenses. Materials often fail because the STAR story is incomplete, ownership is vague, metrics are missing, and there is no adversarial rehearsal.

How it works: from decision to mock defense

guoleme breaks preparation into composable steps:
- Decision framing: the “Laoyan” entry helps decide do / don’t / defer using goals, resource trade-offs, worst-case risk, pseudo-problem checks, and a clear conclusion.
- Evidence mining: resumes, project retros, and reports are mined for projects, ranked by impact, ownership, and fit, then expanded into S/T/A/R with quantified results.
- Mock panel: three to five evaluator personas are generated for the scenario, such as technical, business, skeptical, executive, and HR, with cross-questioning and independent scoring.
- Material diagnostics: decks, scripts, OKRs, and self-introductions are scored for missing sources, vague ownership, logic gaps, weak STAR, and tone issues, with Before/After rewrites.
- Rehearsal and debrief: text or voice rehearsal can simulate time pressure; where captioning or recording is connected, it adds prompt cards and a debrief with score losses and improvements.

Boundaries

Use it when a clear evaluator expects a yes/no decision. Avoid it for plain resume polish, marketing copy, resignation letters, general Q&A, or emotional counseling. It should refine and structure real evidence, not fabricate results; uploaded materials follow the local-processing guidance, and high-stakes moments still benefit from human mock practice.