I have spent years building web applications and commercial platforms, but nothing prepared me for an ordinary Tuesday morning inside a madrasa office in Narsingdi.
I was sitting on a wooden bench across from the madrasa's accountant. On his desk was a stack of five thick, hardbound ledgers with weathered corners and cloth bindings. A mother had walked in to pay her son's monthly tuition and hostel fee. The accountant opened ledger number three, flipped back and forth across dozens of handwritten columns, found the boy's name, wrote out a paper receipt with a ballpoint pen, pressed down hard so the carbon copy underneath would catch the impression, tore out the slip, and handed it over.
The whole interaction took twelve minutes. When the mother asked if her son had been attending his morning classes regularly, the accountant paused, looked over his glasses, and said: "Apa, I only have the payment book here. For attendance, you will have to wait until Huzur finishes his class and ask him directly."
Ten minutes later, a teacher walked in to report that a student had missed three consecutive days. The accountant made a small tick in a pocket diary and admitted he didn't have the guardian's current phone number on hand—he would have to check with the hostel warden after lunch. Right after that, the principal stepped in asking for this month's total fee collection so they could buy kitchen supplies for the residential students. The accountant smiled apologetically: "Give me until Maghrib, Huzur. I need to calculate the rows."
None of these men were lazy or careless. They were deeply committed people working with the only tools they had been given: physical paper, manual tallies, and memory. But watching that scene unfold made one thing unmistakable: the administrative friction in Bangladesh's madrasas is quietly draining enormous amounts of time, energy, and community trust.
The Silent Friction of Paper-Based Institutions
Over the following two months, I visited eleven madrasas across Dhaka, Gazipur, Narayanganj, and Sylhet—both Kowmi and Alia institutions, small village maktabs with fifty students and large residential complexes with over eight hundred boys and girls. The specific routines varied, but the fundamental administrative bottlenecks were identical everywhere:
- Manual Fee Handling & Ledger Disputes: Cash changes hands at an office window. Paper receipts get misplaced by young students on their way home. At the end of every semester, parents and accountants spend hours arguing over whether May's hostel fee was settled or carried over.
- Silent Student Absences: Attendance is marked in paper registers kept in the classroom. When a child fails to arrive in the morning, there is no automatic alert. Parents assume their son is safely in class; teachers assume the boy is at home with family. Days pass before anyone realizes there is a miscommunication.
- Hifz Progress Locked in the Teacher's Mind: In the Quran memorization department, an Ustad listens to fifteen or twenty students reciting individually every single day. A student's exact retention status, recent mistakes, and revision pace live almost entirely in the Ustad's head or in personal jottings that no parent ever gets to see.
- Fragmented Contact Lists: The office has one set of phone numbers written in a register. The hostel warden has another list in a notebook. Individual Ustads save numbers in their personal mobile phones. When an emergency happens, people spend precious time just trying to find who has the right phone number.
- End-of-Term Report Bottlenecks: Compiling semester results means manually transferring marks from individual exam sheets into master grade books, calculating percentages by hand, and writing out physical report cards. It takes two to three weeks of intense staff labor twice a year.
Why Off-the-Shelf School Software Fails in a Madrasa
The immediate question people ask is: "Aren't there dozens of school management software packages on the market? Why not just use one of those?"
Because madrasas are not regular schools with Islamic names stamped on top. Their fundamental operating rhythm is entirely distinct:
A standard school software knows about Grade 1 to Grade 10, semesters, and letter grades like A, B, and C. It has zero concept of Sabak (today's new memorization lesson), Sabki (recent memorization revision), or Manzil / Daur (retention review of earlier Juz). It cannot track progress across 30 Juz, 114 surahs, and 6,236 ayahs. It knows nothing of classical Islamic grading tiers like Mumtaz, Jayyid Jiddan, Jayyid, and Maqbool.
Furthermore, standard software assumes parents have email addresses, know how to create alphanumeric passwords, and can pay invoices via international credit cards. In real-world Bangladesh, rural and semi-urban madrasa parents rely exclusively on bKash, Nagad, and basic phone calls. If software asks an anxious mother to verify her account via an email confirmation link, the adoption rate drops to zero immediately.
The Guiding Principle Behind Al-Maktab
Al-Maktab was conceived not as an academic experiment, but as a direct answer to that Tuesday morning in Narsingdi. Every single subsystem in the platform has been designed around three strict rules:
- Reflect Real Madrasa Workflows: If an Ustad cannot complete his morning student log on a tablet in two taps without typing, the design is wrong.
- Bengali-First, Phone-First for Guardians: A parent must be able to log in with an SMS OTP on their phone, check their child's Hifz wheel and attendance, and pay monthly fees via bKash in under two minutes. No passwords, no English jargon.
- Complete Institutional Integrity: Multi-tenant isolation ensuring that each madrasa's financial records, student files, and staff payroll are completely private, with instant exportable audit trails.
This series of articles documents what it took to design, engineer, and deploy Al-Maktab into production. In the next post, I will share how we designed the teacher interface specifically for Ustads who had never touched a productivity app in their lives.
Mustafa Kamal Hossain
Founder & Principal Engineer at Manfi. Passionate about Laravel, SaaS architecture, and high-performance engineering.