// about
About.
I'm Unaib Amir, a software and enterprise AI consultant based in Islamabad, Pakistan. I've been building software professionally since 2012, working across industries and time zones for clients in Europe, North America, and the Middle East.
It started, like a lot of web careers did back then, with WordPress. Front-end builds, PSD-to-page conversions, then WooCommerce and PHP as the projects got heavier. That led into membership platforms and learning management systems, and from there into the APIs and integration work that makes a multi-product stack actually hold together. Over the years I picked up Python, TypeScript, React, Next.js, Laravel, Django, and Node, plus the infrastructure layer behind them: AWS, Docker, Ansible, Cloudflare. I learn tools as the work calls for them.
The biggest thing I've run was a multi-tenant LMS. 127 organizations and about 125,000 users, all served from one codebase and one Docker image, each tenant isolated on its own database. It touched most of what I care about: the architecture that lets a single deployment scale across a hundred clients, the provisioning to keep them consistent, and the performance problems that only show up at real volume. Identity was a big part of it. I ran an Auth0 estate with one connection per tenant, federated Icelandic eID for one region, and SSO across a family of subdomains. When we had to migrate 120 of those connections to a new setup, I wrote the tooling myself. The whole migration finished in 84 seconds with zero failures.
Infrastructure went the same way. I've managed a fleet of more than 30 servers, and when hosting cost grew past what the business could justify, I led the move from DigitalOcean to Hetzner on my own. It cut the monthly bill by around $5,000 and held 99.97% uptime through the switch. Having done a move like that changes how I weigh infrastructure decisions at the start of a project, instead of waiting until cost becomes the problem.
The AI work is newer, and I try to be straight about what it is. I'm not an AI researcher. I'm a systems engineer who builds AI agents, MCP servers, and automation into real production environments, on work where the AI earns its cost instead of adding complexity. The clearest public piece of it is a free, open-source WordPress plugin called "Agent Abilities for MCP." It gives site operators least-privilege control over what an AI agent is allowed to do inside their WordPress install. It came out of a real problem: agents with broad site access are a risk, and most tooling doesn't give operators a clean way to fence them in. The plugin is the answer I built and put out for anyone to use.
Most clients arrive with one of three problems. A platform that has gotten expensive and slow to maintain, a manual process they want automated, or a product they want built AI-first from scratch. The work looks different each time, but the aim is the same: software that keeps paying for itself well after launch, not a build that looks clean on day one and quietly rots underneath.
I work as a partner, not a pair of hands. I take the outcome you're after, think it through, make the calls on architecture and approach, build it, and tell you plainly what's solid and what still needs watching. I'm most useful on work with real weight: migrations where a wrong move costs you, auth with compliance or scale constraints, and codebases big enough that most people avoid the core. I work with clients worldwide.