Spend · Financial Model
Financial Model
Are you profitable per customer? RepoOps already captures the hardest input to a software COGS model: your AI / token spend, per developer, verb, and model. Add revenue and infrastructure and the Financial Model answers the question every observability and eval competitor leaves on the table: revenue, COGS, gross profit, gross margin, per-unit economics, and break-even.
See it in motion
Where to find it
- Desktop:
http://localhost:4000/, then Spend → Financial Model in the sidebar. - Hosted:
repoops.ai/team/financial-model - Keyboard: ⌘ K, then type “Financial Model”.
- AI-spend source:
GET /api/billing/by-account(localhost) / the hosted AI-spend rollup
What it does for you
Configure
Hosted: the read-only Stripe revenue connector (STRIPE_REVENUE_KEY, a restricted rk_ key) auto-derives MRR / ARR / churn; the Vercel / Neon connectors reuse the same tenant connector vault as telemetry to fill the infra COGS line, with a CSV fallback. All read-only, masked, never logged. Localhost (desktop): a single-user install holds no operator credentials, so the connectors and cross-fleet Benchmarks stay hosted-only and the desktop twin points to repoops.ai for them. On the desktop, revenue and infra are entered by hand and persist to localStorage per repo; the AI / LLM line is still captured automatically.
Use it well
Start on the desktop with your real pricing, paid-seat count, and infra bills, and read the gross-margin KPI. If margin is below your floor, open Scenarios and find the price or scale point that fixes it. When you connect Stripe and your infra providers on the hosted surface, the manual numbers are replaced by live ones and the model updates itself.