The developer left. The platform stayed.

Noqta.tv teaches structured video courses to an Arabic- and English-speaking audience across the Gulf — roughly fifteen courses, sold as five thematic bundles plus one all-access bundle.

The business ran on a custom-coded learning platform built without a long-term architecture. It worked, in the way such systems work: right up until the day it needed to change.

Then the original developer became unavailable. Every new requirement stalled. Arabic students got a worse experience than English students. Years of student records — purchases, enrollments, course progress — sat locked in a database with no clean export path.

And leadership could not answer basic commercial questions. What sold this month. Where buyers came from. Which courses students actually finished. Which paying customers had never opened a lesson.

We moved Noqta.tv to WordPress with Tutor LMS Pro as the learning engine, WPML for bilingual content, and WooCommerce doing one job only — cart and checkout — so course logic and pricing live in one system instead of being duplicated across two.

The client's team had a clear vision for how the platform should look and feel; we brought that design to life on a foundation they can now change themselves — no more paying a developer for every adjustment.

The commercial model was rebuilt around the bundles rather than individual courses. That one decision shaped everything downstream: how access is granted, how the student dashboard is organised, and how revenue is reported.

The platform choice was deliberate. Off-the-shelf, widely supported components mean Noqta.tv is never again dependent on one person's availability — the failure mode that created this engagement was designed out first.

A prior import had brought over the first tranche of students. We moved the full historical student base — every account, every purchase, every enrollment — directly into a live production database.

The approach was deliberately conservative:

• Scripts split into three parts, so no single operation could time out
• Idempotent by design — safe to re-run without creating duplicates
• Scoped strictly to student data; site appearance and plugins untouched
• Pre-import verification queries, with a hard stop if the database state did not match expectations
• Post-import spot checks on individual records, in both languages

Students kept their original passwords, their purchase history, and their place in every course.

We also fixed the product where it was actually breaking: all-access buyers now land on a dashboard organised by bundle in a deliberate learning sequence instead of an unordered wall of courses; completed courses carry a working certificate button wired to the platform's native mechanism; and the Arabic mobile dashboard had its right-to-left overflow bug diagnosed element by element and eliminated.

Noqta.tv now runs on a platform it owns and can extend.

The team went from blind on their own data to fully in control: a dashboard shows what sold, where buyers came from, which courses land, and which paying customers are drifting before they churn. The full historical student base is inside the new system with purchases and progress intact. Arabic and English are genuinely equal — same access, same progress, same certificates. Performance issues that had pushed server response beyond acceptable levels were traced to root causes in the logs and resolved.

Every custom component — the sync layer, the discount integration, the dashboard — is code the client owns, with no recurring licence and no dependency on a developer who might not be reachable next year.

The failure mode that created this engagement was designed out on purpose.