Reports that write themselves every Friday.
Data from everywhere. One beautiful report.
Managers spend Friday afternoons logging into 5+ tools, copying numbers into slides, and formatting reports. It's tedious, error-prone, and repetitive. By the time the report is done, the data is already stale.
Connect your data sources (GA, Stripe, HubSpot, etc.), define your report template, and get a polished report delivered to your inbox every week. AI writes the narrative insights based on the numbers.
Connect sources
GA, Stripe, CRM, etc.
Design template
Drag-drop metrics & charts
Auto-deliver
Email, Slack, or PDF
Create a Next.js app for automated weekly reports. Features: - OAuth integrations: Google Analytics, Stripe, HubSpot - Report builder: drag-drop widgets (metric card, line chart, table) - Preview mode: see report with sample data - Schedule settings: day, time, recipients Use Supabase for auth and storing report configs. Store integration tokens securely with encryption.
Implement data collection: 1. Scheduled job runs at report time 2. For each connected source, fetch this week's data: - GA: sessions, users, top pages, conversions - Stripe: revenue, new customers, churn - HubSpot: new leads, deals closed 3. Calculate week-over-week changes 4. Store snapshot in database for history Handle API errors gracefully—partial report is better than none.
Add AI narrative generation: Send metrics to Claude with prompt: "Generate a 2-paragraph executive summary of this week's business metrics. Highlight: biggest win, biggest concern, recommended action. Be specific with numbers." Include at top of report. Also generate: - One-liner for each metric explaining the change - "Anomaly detected" callouts when data is unusual - Trend analysis comparing to last 4 weeks