Professional services
Multi-tenant CRM for a services group
A group of subsidiaries wanted a single CRM, but each entity needed a hard data boundary for compliance. Shared-table tenancy with a customer_id column was not going to survive an audit.
Backend systems that hold under load.
I design, build and operate production platforms for businesses that cannot afford downtime.
Built with
What I build
Four kinds of work, all of it hands-on. No account manager between you and the person writing the code.
Multi-tenant SaaS platforms
One product serving many customers without their data touching.
- Tenant isolation model
- Per-tenant migrations
- Role and permission design
- Subscription and billing flows
System and API integration
Connecting software that was never designed to talk to anything.
- ERP and SAP interfaces
- Payment provider integration
- Webhook and queue design
- Failure and retry semantics
Platform engineering
Getting the thing deployed, observable and repeatable.
- Kubernetes on AWS EKS
- Helm charts and GitOps delivery
- Infrastructure as code
- Logging, metrics and alerting
Rescue and modernisation
Taking over a system that works but nobody wants to touch.
- Codebase and risk audit
- Framework and PHP upgrades
- Test and pipeline introduction
- Staged migration plan
Selected projects
Client names are under NDA, so each project is described by sector, problem and architecture.
Travel and connectivity
Travel eSIM platform
A paid order has to become a working eSIM, every time, across two systems that can both fail.
- Laravel
- Next.js
- Expo
- Stripe
Retail foreign exchange
Dealer-based FX trading platform
Automated margin and liquidation logic running against live price feeds.
- Laravel
- React
- PostgreSQL
- Redis
Civic technology
National election management platform
A system whose entire year of traffic arrives on one fixed date.
- Laravel Octane
- Swoole
- Horizon
- MariaDB
Consumer healthcare
Patient app for a clinic group
A consumer app that never authenticates against the clinical user table.
- Flutter
- Dart
- Laravel
- PostgreSQL
How the work runs
Find the real constraint
Most briefs describe a solution someone already chose. A short paid discovery establishes what actually limits the system: the data model, the integration, the deployment, or the org chart. Sometimes the answer is that you do not need me.
Ship in slices
Work goes out in vertical slices that run in production, behind flags where needed. You see something real every two weeks, and the integration risk is paid down early rather than discovered in the last month.
Stay on it in production
I run what I build: deployments, monitoring, the incident at an awkward hour. When a handover makes sense, it is a documented one, not a zip file and a phone number.
Ways to work together
Fixed-scope project
A defined system, built and handed over running. Scoped after discovery, priced per phase.
Ongoing engineering
A recurring share of my week as the senior backend and infrastructure person on your product.
Audit or second opinion
A written review of an existing system, its risks and its upgrade path. Usually one to two weeks.
Tell me what is breaking.
Send a short description of the system and the problem. You get a technical answer, not a sales call.
Get in touch
Or email directly[email protected]