BlockTranche: Crypto Payment File Processor

Most service sellers leak 3-5% of every sale to a payment processor — and that's before chargebacks, frozen accounts, and KYC paperwork. CryptoPay is a self-hosted gateway that takes BTC, ETH, and USDT directly into the operator's own wallets, with no middleman in the settlement chain and no customer accounts to manage. The buyer picks a service, sends crypto to a wallet address generated just for that order, and receives a custom fulfillment message the moment the chain confirms.

The problem it solves

Anyone selling consulting, coaching, digital access, or any service that fulfills with a link or instruction text faces the same payment friction. Stripe and PayPal take 2.9% + $0.30 per transaction, demand full KYC of the seller, can freeze funds without warning, and don't natively accept crypto. For a consultant billing $500-2,500 per engagement, that's $15-75 lost to processor fees on every sale before a single hour of work is delivered. For sellers in jurisdictions where Stripe doesn't onboard cleanly — or for those serving a crypto-native customer base who actively want to pay in BTC, ETH, or USDT — the friction is worse: the deal doesn't close at all because there is no rail to settle on.

The alternative paths are also broken. Coinbase Commerce and BitPay charge 1-1.5% and route funds through their own custody before settlement. Off-the-shelf crypto checkout widgets force amount-matching headaches where two buyers send near-identical sums and the operator has to reconcile by hand. None of them ship a fulfillment layer — so the seller still emails a Calendly link or an access code manually after every confirmed payment, and the workflow breaks past about 20 sales a month.

Who needs this most

• Solo consultants and coaches billing $300-3,000 per engagement to crypto-native customers, closing 2-10 deals a month and losing $50-250 of margin per sale to processor fees that don't need to exist.
• Digital product sellers and online educators in jurisdictions where Stripe onboarding stalls (FSU, MENA, parts of LatAm), who can't take card payments at all and need a turnkey way to deliver access codes or course links the moment a buyer pays.
• Web3 service operators — audit shops, dev studios, advisors — whose customers prefer paying in stables and whose current payment story is "DM us a wallet address." That flow costs trust on the buyer side and bookkeeping pain on the seller side.

The moment this hurts: any week a seller closes 3-5 deals over Telegram, manually issues a wallet address per buyer, manually verifies each payment on a block explorer, then manually sends a fulfillment message — and realizes the workflow won't survive a doubling of volume.

The solution — in plain terms

CryptoPay turns a catalog of services into a small, self-hosted storefront. A buyer lands on the store page, picks a service, fills in a name and a contact handle (email or Telegram), and chooses whether to pay in BTC, ETH, or USDT. The system generates a fresh wallet address dedicated to that specific order, shows the exact crypto amount and a QR code, and starts a 30-minute countdown. The buyer pays from any wallet they like. Once the blockchain confirms the transaction, the buyer immediately sees a custom fulfillment message — a Calendly link, an access code, a download URL, a Telegram invite, whatever the seller configured.

Operationally, the backend runs a background monitor loop every 60 seconds that polls BlockCypher (BTC), Etherscan (ETH), and TronGrid (USDT) for incoming transactions to the wallet addresses tied to active orders. As soon as a transaction clears the required confirmation count — 2 for BTC, 12 for ETH, 19 for USDT-TRC20 — the order flips to confirmed and a WebSocket pushes the state to the buyer's open payment page in real time. The admin panel ships revenue dashboards by currency and by service, time-series analytics, a wallet-management view with corporate sweep addresses, and full order history with transaction hashes.

There are no buyer accounts. There is no platform between the seller and the funds. Settlement is on-chain.

Value delivered — what you get

• Keeps 100% of every payment — replacing a typical 2.9% + $0.30 Stripe fee on a $499 consulting sale saves $14.77 per transaction. At 10 sales a month that is roughly $1,800 of margin recovered each year, with the operator paying only blockchain network fees.
• Closes deals that Stripe blocks outright — sellers in onboarding-restricted jurisdictions get a working settlement rail today, not a six-month appeal queue.
• Cuts post-sale fulfillment from a manual 5-minute task to zero — the per-service fulfillment text fires automatically the moment the chain confirms; no more sending Calendly links by hand after each payment.
• Eliminates amount-matching reconciliation — every order gets its own wallet address, so two buyers sending identical amounts never collide in the seller's books.
• No chargebacks, no frozen accounts, no surprise compliance holds — on-chain payment is final, custody is the seller's, no third party can reverse or freeze the funds.
• Three currencies inside one container set — BTC, ETH, and USDT-TRC20 all run from one admin panel and one database, instead of three vendor accounts to reconcile.
• Real-time buyer reassurance — the WebSocket-driven payment page moves from "waiting" through "detected" to "confirmed" without the buyer reloading or opening a block explorer.

Where it delivers outsized value

This lands hardest in three contexts. First, solo operators and small studios serving a crypto-native customer base — consultants, coaches, audit shops, dev houses where buyers prefer stables and the seller is also the bookkeeper. Second, digital-product and access-based businesses in jurisdictions where Stripe and PayPal don't onboard the seller cleanly; CryptoPay gives them a same-day payment rail with no application paperwork. Third, sellers who want full sovereignty over their funds — no platform custody, no "we noticed unusual activity" account freezes, no per-transaction fee skimmed by an intermediary. The common thread: small to mid-volume operators whose unit economics suffer disproportionately from 3% processor fees, whose product is digital enough to fulfill on a single confirmed payment, and whose customers are already comfortable holding crypto.

Distinctive features — why this over the alternatives

• One wallet address per order — no amount-matching, no reconciliation, no confused buyers paying to a shared address. Each order's address ties cleanly to one buyer and one fulfillment.
• Custom per-service fulfillment text — the field that turns a payment tool into a delivery tool. Each service carries its own freeform message (instructions, Calendly link, access code, channel invite) surfaced the moment the chain confirms.
• Three-chain coverage in one self-hosted stack — BTC via BlockCypher, ETH via Etherscan, USDT-TRC20 via TronGrid. Adding a fourth chain (Solana, Polygon, Base) is a new monitor class slotted into the existing scheduler, not a new product.
• Encrypted private-key custody under the operator's key — Fernet-encrypted keys live in the seller's own Postgres, decrypted only when sweeping funds to a corporate wallet. No third-party custody, no exchange-account dependency.
• Token-based URLs end-to-end — services and orders are addressed by URL-safe random tokens, not by names or sequential IDs. No price scraping, no order enumeration, no leak of the service catalog through the URL bar.
• Single-file config.ini as the pilot cabin — ports, domains, API keys, wallet addresses, and confirmation counts all live in one operator-facing file. The runtime .env is regenerated from it on every start, so configuration drift stays close to zero.

Under the hood — built to last

The backend runs on FastAPI with SQLAlchemy 2.0 async and Alembic migrations against PostgreSQL 15 — boring, well-supported foundations that will still work in five years. The frontend is Next.js 15 with React 19 and TailwindCSS, rendered as a standalone container. Wallet generation uses eth_account and tronpy for local key derivation and BlockCypher's address API for BTC; private keys are encrypted with Fernet before they touch the database. The whole stack ships as three Docker containers (Postgres, backend, frontend) and stands up on a single $10-20/month VPS with no SaaS dependency for any critical path. Self-hostable end-to-end, container-native, no cloud lock-in.

Current maturity

The build is feature-complete against a documented 5-phase implementation plan: backend foundation, blockchain monitors, admin panel, frontend storefront, and DevOps tooling. All three blockchain monitors are implemented and the admin panel covers dashboard, orders, services, wallets, and analytics. The repository carries roughly 6,800 lines of code across the Python backend (~2,500 lines) and TypeScript frontend (~2,500 lines), plus an architecture document, concept document, operator guide, and a set of shell scripts (setup, start, down, cleanup, monitor) wired around the same config.ini. Last activity 2025-03-08. Honest framing: the build runs in development against mainnet APIs and is ready for a live VPS deployment behind a reverse proxy, but it has not yet been operated against production traffic — the next step is a deployed pilot with a real seller and real revenue.

Roadmap — what's next

The immediate next milestone is that live pilot — validating the unit-economics promise of zero processor fees against actual transaction volume from a real consulting or digital-product seller. From there, three concrete extensions follow: automated fund sweeping from per-order addresses to corporate wallets on a scheduled cadence (the encrypted-key plumbing is already in place; what's missing is the scheduler), additional chain support starting with USDC on Solana and Base since those are the rails crypto-native buyers most often request, and a buyer-side receipt option that delivers the fulfillment message via email or Telegram for sellers operating in channels the buyer already uses.

The longer arc is a multi-tenant edition that lets a small group of consultants share a hosted instance under one operator's umbrella — opening a low-friction "pay-for-tenancy" engagement model alongside the existing "build-it-yourself" mode.

Working with the architect

CryptoPay is available in three engagement modes. A consultant, agency, or studio that wants the storefront live under its own brand can commission a custom build modeled on this codebase, tuned to a specific service catalog and look. An existing platform with a checkout flow can integrate the payment-and-fulfillment layer into its own stack — wallet generation, blockchain monitoring, and confirmed-payment hooks lift out as a self-contained subsystem. And operators planning their own crypto-payment build can engage in strategic advisory on architecture, multi-chain coverage strategy, and the operational realities of running a self-custodial settlement layer in production. Reach out via sintegrium.io or LinkedIn for a 30-minute scoping call.

---

Built by Yurii Staryk · Solution Ecosystem Architect