Blockchain Loyalty Platform
Core engineer at qiibee, building a multi-tenant, blockchain-backed loyalty platform for 20+ brand partners on Elixir, Phoenix, and AWS Lambda — points and NFTs redeemable for real-world value.
qiibee (Zürich, Switzerland) runs a multi-tenant, white-label loyalty platform on its own blockchain. Each brand partner gets an isolated loyalty program; members earn points and NFTs and redeem them for things with real value — airline miles, crypto, fiat, and gift cards. Because those balances map to money, the on-chain layer is the system of record, not decoration, and correctness is non-negotiable. I was one of two primary backend engineers across a large Elixir umbrella of 15+ services, plus the serverless and blockchain-signing repositories.
What I built:
- The redemption engine. The layer that converts loyalty points into real-world value — a pluggable provider-behaviour pattern in Elixir that let us add new exchange partners without touching the core. Behind it I integrated Coinbase and Binance (including just-in-time crypto purchases at redemption time), PayPal for fiat, and airline mileage programs, each with its own auth, quirks, and failure modes.
- The Shopify integration, as primary engineer. Coupon and price-rule management, paying with loyalty miles at checkout, merchant-credit billing, member search, and a lot of webhook processing to keep both systems in sync — opening a new e-commerce customer segment for the company.
- Data-integrity recovery after a blockchain outage. Authored the NFT-balance reconciliation and on-chain nonce re-sync tooling that converged on-chain state back to the database ledger after the chain was inaccessible for hours while transactions kept arriving (details below).
- Core blockchain signing-service capabilities. The bridge between the platform and the chain: token-contract deployment, NFT minting, transfer and sale, batch balance reads, and chain mirroring; later led the service’s Node.js runtime upgrade.
- The serverless layer. A primary author of a suite of AWS Lambda functions (Serverless Framework + TypeScript) — a transaction polling/reconciliation service, automated email notifications, and scheduled data jobs.
- A new Hotels product vertical. Merchant and end-user apps with QR/email reward flows and PayPal-based payments.
- Security hardening. Patched multiple data-leak vulnerabilities in token-exchange endpoints, added cryptographic wallet-ownership (signature) verification, and tightened access control on data-export functions.
The outage reconciliation (what I’m most proud of): Every points and NFT movement was recorded in a Postgres ledger and then executed on the blockchain. During a multi-hour window the chain and its signing service became inaccessible, but members kept earning, exchanging, and redeeming. When the chain returned, the database and on-chain state had diverged and nonces were stale — and we couldn’t blindly replay everything without risking double-credits. Treating the database ledger as the source of truth, I built tooling that derived each user’s expected NFT holdings from the ledger and diffed it against what the chain actually reported, re-crediting shortfalls and flagging excess, plus a nonce re-sync that was the prerequisite for safely replaying transactions. It was idempotent and dry-run-first by design. The result: 500+ affected transactions reconciled and every balance made whole, with no customer-visible loss.
Impact:
- Built a sales-automation tool that lifted the team’s prospecting throughput ~3x (Elixir + JavaScript), and joined direct product pitches to the hospitality sector.
- Reduced deployment time by 50% for the Elixir backend and improved data-retrieval performance by 1400% within Lambda constraints.
My role: Backend engineer (Aug 2021 – Oct 2024), working across Elixir/Phoenix, Node.js/TypeScript, and AWS serverless, plus the blockchain signing service.