I'm Muhammad Muneeb, a Laravel backend developer. I build payment systems that hold up in production and put AI to real work inside them. Stripe, PayPal, and Connect on the money side, Gemini and RAG on the AI side.
public function handle(Request $request): Response
{
$event = $this->verifySignature($request);
// Insert first. Stripe can deliver the same event twice.
try {
WebhookEvent::create([
'event_id' => $event->id,
'type' => $event->type,
]);
} catch (UniqueConstraintViolationException) {
// Already handled. Ack so Stripe stops retrying.
return response()->noContent();
}
ProcessStripeEvent::dispatch($event->id);
return response()->noContent();
}
Stripe, Stripe Connect, PayPal, Xero, Zoho, QuickBooks
Five years with Laravel, two with Next.js. Payment flows are where bugs are most expensive, and I've learned the edge cases the hard way. These days I also build AI features that do real work inside Laravel apps, the kind that need real engineering to run safely in production, not a chatbot bolted onto a sidebar.
Payment gateway integration has been part of my work at four of these five companies, going back to 2020.
Get Licensed
Coderatory
HulHub
LathranSoft
Herobotic
Virtual University of Pakistan
Jan 2023 to present
Aptech Learning Pakistan
Jan 2019 to Jan 2021
Five systems built from scratch, each one live to click through and test yourself.
Invoicing where the payments and the bookkeeping are one system, not two.
A multi-tenant invoicing platform tying Stripe, PayPal, and Xero into one system across four roles (owner, staff, customer, admin), so the money coming in and the books staying straight are one system, not two. Gemini categorizes every transaction, flags anomalies like duplicate charges, and drafts payment reminders a human sends in one click.
Answers drawn only from your own documents, with citations and no invented sources.
A multi-tenant RAG platform where each workspace uploads its own documents and gets answers drawn only from those files, with inline citations and no hallucinated sources. Gemini handles embeddings and streaming chat. Tenant isolation is enforced in three separate layers, so one company's data can never surface in another's results. Queued document processing and cosine-similarity retrieval with no external vector service.
Upload fifty resumes, get a ranked shortlist with a reason for each.
Upload fifty resumes, get a ranked shortlist. Each resume is unpacked, queued as its own background job, parsed into a structured profile, and scored 0 to 100 against the job description with a one-line reason. Recruiters work a sorted review queue instead of a pile, then export the shortlist as CSV or fire a webhook. Duplicate detection and a full audit trail on every decision.
Sellers onboarded, payments split automatically, payouts explained in plain language.
Seller onboarding, automatic split payments between platform and sellers, automated payouts, and AI-generated payout summaries so each seller sees exactly what they earned and why.
Recurring billing that chases a failed payment instead of losing the customer.
A Laravel Cashier and Stripe subscription system that handles what most setups skip: failed payments. Plan management and checkout, then when a renewal fails it fires AI-written dunning emails that chase the customer instead of letting them churn silently.
Most of this work comes down to one thing: money moving correctly, and staying correct when something fails.
Taking payments properly the first time: checkout, refunds, disputes, and the failure paths most integrations skip.
For teams adding payments, or replacing an integration that keeps breaking.
Typical deliverables
Plans, upgrades, downgrades, proration, and the failed-payment recovery that decides whether churn is silent or fixable.
For SaaS products billing on a recurring cycle.
Typical deliverables
Seller onboarding, split payments between platform and sellers, and payouts that reconcile without manual work.
For marketplaces paying out to third-party sellers.
Typical deliverables
Wiring payment data into Xero, Zoho, or QuickBooks so the books match the bank without anyone rekeying invoices.
For businesses reconciling payments by hand every month.
Typical deliverables
Document intelligence, structured extraction, categorization, and anomaly detection, built with validation and queued processing so a bad model response never reaches your database.
For teams who want AI doing real work, not a chatbot in a sidebar.
Typical deliverables
Idempotency, retries, and dead-letter queues, plus debugging the duplicate charges an unreliable endpoint has already caused.
For anyone who has seen a customer charged twice.
Typical deliverables
PHP 5.x to 8, CodeIgniter or .NET to Laravel, done incrementally so the thing that currently earns money keeps working.
For teams stuck on a codebase that is expensive to change.
Typical deliverables
A call to understand the money flow: who pays, who gets paid, what happens when it fails. I ask the awkward questions early.
A written scope with the edge cases listed, so we agree on what done means before I write code.
Built in reviewable pieces. You see working software as it lands, not a black box for six weeks.
Tests on every payment path, plus the failure cases: declined cards, duplicate webhooks, timeouts mid-transaction.
Deployed with a rollback plan and live webhook monitoring, so a problem shows up on a dashboard and not in your inbox.
I stay available after launch. Payment bugs surface in production, and I do not hand over and disappear.
Whether it's a payment system that has to be right or an AI feature that has to run safely in production, tell me what you're building. I reply within a few hours during EU/US business-hours overlap.
The more detail you give me, the more useful my first reply is.
Based in Karachi, Pakistan, working with EU and US clients.