Skip to main content
Status: Proposed (RECONSTRUCTION — requires validation) · Date: 2026-08-17 (reconstructed) Refs: ../constitution/founding-constitution.md
⚠️ Reconstructed from the material we have, not recovered. Tags: [derived] is supported by evidence in the repository · [inferred] is deduced · [proposed] is a gap I filled.

Context

Selling from Chile to Chilean and international customers creates two different problems. Locally, an invoice is a legally structured document that must be issued to the SII, and card payments are better served by local rails. Internationally, sales tax and VAT across dozens of jurisdictions is a compliance burden that a merchant of record absorbs entirely. [inferred]

Decision

Fintoc for Chile, Paddle as merchant of record internationally, LibreDTE for SII documents. [derived] Both payment providers sit behind one PaymentPort [inferred], and the rating engine is ours (ADR-016): we compute what is owed from usage snapshots, and the provider only executes the charge. [derived] That split matters — a provider that computed our prices would make our pricing model its constraint. Fintoc charges a card the tenant enrolled, for an amount we calculate. Paddle handles usage-based billing and absorbs international tax. LibreDTE issues the Chilean tax document with variable lines. [derived]

Consequences

  • International tax compliance is bought, not built.
  • Chilean invoicing is correct against the SII from the first invoice.
  • Our pricing model is ours: usage-based, published unit prices, spend caps. − Two payment integrations, two reconciliations, two failure modes. − A tenant’s geography determines their provider, which the billing model has to carry.

What I could not determine

Whether Paddle was chosen over Lemon Squeezy or Stripe for a specific reason worth recording. A merchant-of-record decision usually has one. [proposed]