R.Song Studios — ops dashboard for a 3-location fitness studio

R.Song Studios runs pole and aerial fitness across three Calgary locations. Their operational data lived in scheduling-software exports, bank statements, and an Airtable payroll workflow. I built them a single dashboard that tells the story of the business every month.

What it does

• Sales analysis — month-over-month bridges showing exactly where revenue moved and why
• Cohort retention heatmaps — which intake months stick, which classes retain
• Instructor profitability — revenue per instructor against class costs
• Payroll & pay runs — replaced the manual Airtable flow with pay-run generation and statement PDFs
• AI finance import — bank statements go in, DeepSeek categorizes transactions, the books stay current
• Win-back drafts — lapsed-member emails drafted by AI for the owner to approve and send
• Team management — RBAC with a full audit log, so staff see what they need and nothing else

Built like production, because it is

This is a client’s business, not a demo: Sentry error tracking, UptimeRobot monitoring, GitHub Actions auto-deploy, and a Supabase schema with 24 migrations and 50+ views doing the analytical heavy lifting in the database rather than in application code. It runs on Next.js 16 deployed to Cloudflare Workers — fast, and the whole thing costs a rounding error to host.

The data model and payroll engine from this build became the foundation for a multi-tenant operations platform I’m now developing, with R.Song as tenant #1.