API & backend development services Backends built to be
operated, not just shipped
Faizan Amir builds operable Node.js, Express, PostgreSQL and MongoDB backends and has built 30+ REST and GraphQL APIs across client products, including a live telehealth platform that reduced approval turnaround by 45% and despatch time by 85%, with payments reconciled against Stripe.
What this includes
Orders, prescriptions, subscriptions and refunds use explicit states and transitions. Volumize connects intake, clinical review and fulfilment in one traceable lifecycle.
Provider-backed entitlement, webhook reconciliation, idempotency and monotonic status prevent duplicate or late events from corrupting orders.
Signed S3 uploads and edge delivery keep image-heavy workloads away from application request workers.
Expand-and-contract migrations let schemas evolve while live products keep serving users.
Phased dependency remediation cleared every high and moderate finding across Volumize’s 1,127-package graph.
How the work runs
- *[01]
Model the domain
Create one definition per entity and share it across every application that uses it.
- *[02]
Draw the state machine
Define statuses, allowed transitions and irreversible actions, including refunds and cancellations.
- *[03]
Build the endpoints
Implement validated REST or GraphQL endpoints with authorisation next to data access.
- *[04]
Guard the money
Add signature, ownership and amount checks, idempotency, replay protection and settlement tests.
- *[05]
Hand over something operable
Ship CI/CD, logs and metrics so releases and production diagnosis are routine.
Where this has been built
Volumize
Telehealth E-Commerce Platform
Live telehealth commerce with 250+ subscriptions and £31,000+ revenue: approval turnaround down 45%, despatch time down 85%, and 15 admin hours a week returned.
READ THE VOLUMIZE CASE STUDY →WisdomUp
Multi-Service E-Commerce Platform
* IN DEVELOPMENT
Eight payment guard layers, from signature verification to monotonic order state, plus 14 tests on settlement confirmation.
READ THE WISDOMUP CASE STUDY →Carder
AI Mini-Site Builder
Zero-downtime schema evolution, reconciled Stripe subscriptions and signed S3 media delivery for a live product.
READ THE CARDER CASE STUDY →SeeTrack
Field-Force & Distribution Platform
Ten modules and 334 API endpoints of a live field-force platform: a stock ledger where one guarded function moves every piece, a permission matrix whose parity test caught 119 access errors in its first seed, before release, and double-entry finance with ten nightly control checks, in pre-launch.
READ THE SEETRACK CASE STUDY →
[STACK]
- Node.js
- Express
- TypeScript
- PostgreSQL
- MongoDB
- Mongoose
- GraphQL
- Stripe
- AWS S3
- Socket.IO
Questions about API & backend development
REST suits most product APIs. GraphQL is useful when clients need different views across a broad data graph. In either case, validate at the boundary and authorise where data is accessed.
Verify signatures, ownership and amounts; deduplicate transactions; and reject events that would move an order backwards. Immediate confirmation plus webhook reconciliation prevents a closed browser from losing a paid order.
Yes. The first step is a written model of the existing schema, followed by expand-and-contract migrations that avoid maintenance windows.