Training Completion Dashboard

The problem this kills

Every leadership meeting, someone asks "what's our training completion rate?" - and three people pull up three spreadsheets with three different numbers. One exported from the LMS last Tuesday. One a manager keeps by hand. One that double-counts people who took the same course in two systems. Nobody trusts any of them, so the conversation becomes about the data instead of about the people who are overdue.

The root problem isn't effort - it's that completion records live in different tools, with different names for the same person and the same course, and nobody has agreed what "complete" even means. Counting? Including or excluding terminated staff? Overdue from the due date or a grace period?

This plan builds the single, dated, agreed-upon view that ends the argument.

What you'll build

A web dashboard, login-protected, that:

• Pulls completion records from one or more sources (your LMS export, other tools, or a Google Sheet) and reconciles them - deduping the same person + course + period across systems.
• Maps every record to your org structure so you can slice by department, location, course, and manager.
• Computes completion rate, overdue count, and trend over time for each cut, against targets you define.
• Lets a leader drill from any number down to the actual list of people behind it.
• Restricts managers (row-level security) so they only ever see their own organization's data.
• Requires the L&D lead to approve the data sources and metric definitions, then publishes a dated snapshot - so leadership always knows exactly what they're looking at and as of when.
• Exports the rollup as a clean CSV.

What's inside the Implementation Plan

The plan opens by interviewing you about your business - your tools, the real names and codes in your data, your reporting structure, your typical and peak volumes, and how you define "complete" and "overdue." It reads a short tailored spec back to you and waits for your thumbs-up. Everything it builds after that - the data model, the dedupe rules, the metrics, the org cuts - is shaped to your answers, not a generic template.

Inside you'll find:

• The full discovery interview (8-12 plan-specific questions).
• A data model for completion records, the org/reporting tree, courses, and metric definitions.
• The reconciliation + dedupe logic (person + program + period across sources).
• The metric engine: rates, overdue counts, and trend lines by every cut.
• The drill-down screens and the manager-scoped RLS.
• The approval gate and dated snapshot publishing.
• The CSV import fallback and CSV rollup export.
• Step-by-step build prompts you paste straight into your AI coding agent.

The governance it includes (this is the point)

• Login so only your team can open the dashboard.
• Row-level security so managers drill only into their own org - never another department's people.
• A full audit trail of who loaded what, who changed a metric definition, and who published each snapshot.
• A human approval gate: the L&D lead reviews and approves the data sources and metric definitions before any snapshot is published to leadership - the AI drafts the rollup, a person signs off, and only then does it go live.
• Duplicate guards so the same completion record can't be counted twice across systems.

Who it's for

L&D leaders and executives who keep getting conflicting completion numbers from different spreadsheets and want one trustworthy, dated view - plus the ops/BPM person who's tired of being the human reconciliation engine.

You've got this - paste the first prompt and let the interview tailor the build to your business.