How WeActually Build This
It started with one developer and state-of-the-art AI. Five months in, the foundation is built — and the team is just getting started. Here are the real numbers, real tools, and real methodology behind Tawaasal.
The Numbers
Real metrics from our git repositories. No vanity numbers — every stat is pulled directly from our codebase.
438,000+
Lines of Code
1,400+
Git Commits
65
Feature Specs
4,200+
Tracked Tasks
32
Django Apps
24
Mobile Features
1,000+
i18n Files
5
Months Active
Spec-Driven Development
Every feature starts as a specification, not a Jira ticket. We built a custom framework called SpecKit that enforces a rigorous pipeline from idea to production code.
Specify
Natural language feature description becomes a structured spec.md with user stories, acceptance criteria, and scope boundaries.
Plan
Technical architecture, data models, API contracts, and file structure — all documented before a single line of code is written.
Tasks
Dependency-ordered task breakdown with parallel markers, validation checkpoints, and phase gates. Every task has file-level traceability.
Implement
AI executes tasks sequentially, loading context per-task from the constitution and pattern references. Micro-reviews after each task.
Review
Constitution-aware code review with severity levels (BLOCKER/ERROR/WARNING/INFO), spec traceability, and automated fix suggestions.
We have written 65 feature specifications across the project, generating over 4,200 tracked tasks with 68% completion rate.
github/spec-kitThe Constitution
A 2,168-line living document of 24 engineering principles that every line of code must comply with. From UUIDv7 primary keys to tenant isolation, i18n parity to accessibility standards — the constitution is the single source of truth.
24 principles covering code quality, security, permissions, i18n, accessibility, performance, and design consistency.
The Tools We Use
Building a production app with AI requires more than one tool. Each serves a different purpose in the development workflow.
The Full Stack
Three platforms, one codebase philosophy. Every layer follows the same constitution and spec-driven process.
Backend
v3.14 / v5.2Python 3.14, Django 5.2, DRF — 32 apps, 509 files, 151K lines. Multi-tenant via django-tenants with PostgreSQL 17. UUIDv7 primary keys, centralized permissions, audit logging.
Frontend
v19 / v5.9React 19, TypeScript 5.x, Vite — 629 files, 183K lines. Tailwind CSS, Framer Motion, Zustand state, react-i18next with full EN/AR bilingual support and RTL layout.
Shared i18n
1,003 files1,003 translation files across English and Arabic. Every user-facing string goes through the shared i18n pipeline — backend, frontend, and mobile all consume the same source of truth.
Chapter One
Five months from first commit to a full-stack production platform with 438K+ lines of code. This is the foundation — what comes next is bigger.
November 2025
Project kickoff. Backend foundation, multi-tenant architecture, user management, and the first Django apps. SpecKit workflow established.
December 2025
Core features: homework, grading, attendance, calendar, messaging, permissions system, and subscription management. 20+ specs completed.
January 2026
Advanced features: safeguarding, communication templates, data import/export, rich text editor, student lifecycle management.
February 2026
Flutter mobile app: student, guardian, teacher, and staff experiences. Offline sync, theme system, and cross-platform parity.
March 2026
Marketing website, CMS, public pages, and this very page you're reading. UI design revolution across web and mobile.
What's Next
Growing the team, onboarding pilot schools, and building the features that turn a platform into a product. The architecture is ready — now it scales.
Want to See It in Action?
Five months of engineering by one developer with AI. Imagine what a growing team will do. Try the demo and see what's already possible.