Senior backend engineer: event-driven data pipelines, PostgreSQL and reliability for payment and blockchain systems.
I've been shipping production software for 8 years. I own systems end to end: architecture and ADRs, schema and access control, CI/CD and Kubernetes, load testing and failure handling. Based in Abuja, Nigeria, working remotely.
Heightproof: catches the EVM RPC failures that return 200 OK
Your RPC provider answered in 40 ms with 200 OK, and the answer came from a node 40 blocks behind. Heightproof is a self-hosted platform that detects this. It uses block-pinned cross-provider verification, plus stale-head, reorg-lag, silent-failure and throttling detectors, and RPC cost accounting.
- Pipeline: Node SDKs → Fastify ingestion → Redis Streams → worker → PostgreSQL → React dashboard (live over WebSocket)
- Measured: 59,000 events accepted in 30 s at 2.9 ms p50 / 37.8 ms p99, and ~1,030 events/s per worker (performance)
- Production delivery: Helm chart, Kubernetes CI/CD with kind e2e tests, CodeQL, ADRs
| Project | What it is | Stack |
|---|---|---|
| Inkkrystal (client) | Reading and publishing platform for web, Android and iOS: row-level security across four roles, royalties and wallets, and server-side Google Play and Apple purchase verification | TypeScript, Supabase, PostgreSQL, React, Capacitor |
| Xeinon | Real-time token discovery on Base: WebSocket detection, backfill and bytecode scanning into a provenance-tracked PostgreSQL index | TypeScript, Supabase, PostgreSQL, React |
Merged contributions to:
Core: TypeScript · Node.js · PostgreSQL · Redis · SQL Infra: Docker · Kubernetes · Helm · GitHub Actions · AWS · Linux Also: React · Supabase · PHP · EVM / JSON-RPC
macjayz90s@gmail.com. I'm open to senior backend and infrastructure roles.