Solutions

Where we think Orynt could help.

Four target use cases, each written as a workflow in terms of features that exist in the app today. They are hypotheses: no organisation uses Orynt, and we are not describing any customer.

  • Target use cases
  • No customers
  • Regtest only

Target use case No customers

Developer payouts

Who it might suit
A product team or open-source maintainer that pays people from code: bounties, bug rewards, creator or referral rewards, sent over Lightning when something happens in their own system.
Workflow
  1. Create from codeThe backend calls POST /v1/obligations (or obligations.create in the SDK) with an API key. The payee, exact amount and invoice are recorded and the invoice is checked.
  2. Humans approveThe payout waits for the policy's signed approvals. API keys cannot approve, so software cannot release its own spending.
  3. Execute onceThe backend calls execute with an Idempotency-Key. A 202 means “not certain yet”: it waits or reconciles and never sends again.
  4. Close the loopA signed payment.settled webhook updates your own records, and the receipt can be attached to the reward.
Plan fit
The free Developer plan allows 25 payouts a month, 2 API keys and 1 webhook endpoint.
What we don’t know
Lightning invoices are single-use, so a payee supplies a fresh invoice for every payout, or a Lightning Address that Orynt resolves into a checked invoice. Whether that friction is acceptable for rewards is something we have not tested.

Target use case No customers

Contributor funding

Who it might suit
Open-source projects, community programmes and grant makers that pay contributors in sats, on a schedule or in rounds.
Workflow
  1. Round of payoutsA maintainer uploads a CSV (counterparty,amount_sats,invoice, optionally memo and external_ref, up to 200 rows). A preview flags bad rows without creating anything. Batch CSV needs the Startup plan or above.
  2. Standing stipendsA weekly or monthly schedule opens an “invoice needed” task each cycle. It never pays by itself: the contributor supplies a fresh invoice, which becomes a normal draft payout. Schedules need the Growth plan or above.
  3. Approve and sendApprovers sign each payout. Executing a batch sends every approved item through the normal single-use path, one by one; one failure or doubt never hides the others.
  4. Report to fundersA payout report and per-recipient statements export as CSV. They sum real settlements only and carry no fiat values.
Plan fit
Batch CSV starts at Startup; recurring schedules start at Growth. On Developer you can still create payouts one at a time.
What we don’t know
Many such projects have little budget. This may suit the free plan and a self-hosted setup better than anything paid. A receipt proves a payment, not that a milestone was met.

Target use case No customers

Contractor payments

Who it might suit
A small team that pays freelancers or contractors in Bitcoin over Lightning and wants a second person to approve, and a record of who agreed to what.
Workflow
  1. Know the payeeEach contractor is a payee record, optionally bound to their Lightning node key so that a swapped invoice is caught. A payee can be suspended.
  2. Record the workA payout carries a memo and your own reference, such as an invoice number, next to the amount and invoice.
  3. Maker-checkerA policy can require that the person who creates a payout is not the one who approves it. Each approver signs the exact terms.
  4. Proof for both sidesThe contractor can verify the receipt offline with the standalone verifier, without an Orynt account. A recipient statement lists what was paid to them.
Plan fit
A small team fits within the Startup plan's 5 members and 250 payouts a month; the Developer plan allows 3 members.
What we don’t know
Orynt does not produce invoices, tax documents or fiat accounting, and does not check that work was delivered. Contractors have to be comfortable producing Lightning invoices.

Target use case No customers

Merchant disbursements

Who it might suit
A merchant, marketplace or affiliate programme that owes refunds, supplier payments or commissions in Bitcoin, event by event or in a regular run.
Workflow
  1. Collect invoicesRecipients send Lightning invoices. The batch preview checks each one: if an invoice carries an amount it must equal the amount in the file, and the same invoice twice in a file is flagged.
  2. Review and submitValid rows become draft payouts in a batch. Submitting the batch asks for approval on every item; failed rows are listed with the reason.
  3. Run the batchEach approved item is executed on its own. A batch is “completed” only when every item has settled; it shows “partial” otherwise.
  4. Handle exceptionsAnything uncertain shows as “Checking wallet record” and opens an incident with notes and an owner, instead of being retried blindly.
Plan fit
Batches need the Startup plan or above (250 payouts a month); a batch that would exceed the monthly limit is refused whole.
What we don’t know
This is the weakest fit we can see: low volume, hard to reach, and other tools may cover it. A refund is a new payment, never a reversal, and recipients must provide an invoice or a Lightning Address. Lightning Address is verified only against a local test server, not real providers.

Plan limits are read from the current plan catalogue; prices are hypotheses under test.

Probably not a fit.

  • You want something to hold, convert or lend funds, or to run KYC. Orynt does none of that.
  • You want fully automatic spending with no human approval.
  • You need an audit report or certification today. None exists.
  • You need mainnet payments now. Mainnet is disabled.
  • You need a wallet other than LND. NWC, BTCPay and Core Lightning adapters are not built (integrations).
  • You make only a handful of payouts a month. A wallet and a notebook may serve you well.
  • You already use a tool that does the approvals you need. If BTCPay Server’s payout approvals are enough, use them.

Does one of these sound like you?

Try the flow on regtest with a free account, or tell us how you pay people today, including what goes wrong. That is the most useful thing we can learn right now.