Work · Own product · Waitlist
Mithaqq
A privacy-first matrimonial platform where the trust model is the architecture.
Mithaqq is a curated Islamic marriage platform built around a constraint most social products never face: profiles must stay invisible by default, and every conversation is accountable. The product requirements are enforced in the data model, not in the UI.
The problem
Matrimonial apps borrow their mechanics from dating apps: open browsing, public profiles, unmoderated messaging. For users who want marriage with intention — and for the families involved — those defaults are not a minor mismatch, they are disqualifying. The trust requirements are the product, so they cannot be bolted on afterwards.
How we built it
Consent as a data-model constraint
Profiles are visible only on mutual consent. Enforcing that at the query layer rather than the view layer is the difference between a privacy promise and a privacy property — it means no future feature, endpoint, or admin screen can accidentally leak a profile that was never meant to be browsable.
Manual review as a designed workflow
A curated community means humans in the loop, which means the review queue is a first-class system with its own state machine, throughput limits, and audit trail — not an inbox someone checks. Designing for the reviewer early is what keeps curation from collapsing as volume grows.
Supervised conversation and wali involvement
Guardian participation is a genuine third role in the conversation model, not a notification setting. Supporting it properly means the messaging system has to represent supervised threads natively, with the permissions and visibility rules that implies.
Decisions worth defending
- Waitlist before launch, deliberately
- A curated community with weak early moderation is worse than no community, because the first cohort sets the norms. The waitlist exists to make the first intake reviewable at a pace humans can actually sustain.
- Collect less, on purpose
- Every additional field is a liability in a product handling matrimonial data. The schema is deliberately narrow — the cheapest way to protect sensitive data is not to hold it.
What shipped
- In private waitlist ahead of a curated first intake
- Consent and visibility rules enforced at the data layer
- Moderation and guardian-supervised messaging designed in from the start
The service behind this
Let's talk
We take on a small number of projects. Tell us about yours.
Reply within 1 business day · NDAs signed before details
- Founded · Cairo
- 2022
- Products built
- 2
- End to end
- Senior
- Scope → estimate
- 1wk