Most tools collect ratings. Meridian's whole point is the gap between what you think and what your manager thinks — tracked monthly, visualised clearly, and surfaced before it becomes a problem. Persistent gaps >1pt for 2+ months trigger an automatic alert.

KRAs live in Excel — where your HR team already manages them. Upload once, Meridian parses the full hierarchy including the dedicated quarterly sheet. All ratings persist in Supabase. No new workflows for the team, full data integrity for HR.

Every AI feature — insights, SWOT coaching, roadmap, 1-on-1 prep — reads your actual ratings, your actual gaps, and your actual goals before generating anything. Claude doesn't give generic advice. It gives advice specific to your month, your KRAs, and your numbers.

The entire platform — both dashboards, all charts, all AI, all data — is one HTML file deployed on Netlify. Supabase handles auth and storage. Anthropic handles AI. Nothing to maintain, nothing to scale, nothing to break. Deploy by dropping a file.

Add to home screen on any phone. Service worker caches the app for offline access. Bottom navigation collapses the full sidebar into a scrollable tab bar. Inputs are sized to prevent iOS zoom. It works like a native app because it was designed to be one.