Approach
Careful engineering, honest AI.
How we build, how we use AI, and how we handle sacred text and your data.
How we build
We keep scope small and finish what we start. Every release is tested and measured before it ships, and we keep maintaining the code after launch.
Arabic and right-to-left are part of the first commit, not a translation pass at the end. Layout, typography, search and dates are designed for Arabic from the start.
How we use AI
We use AI where it saves people time without taking their judgment: drafting, transcription, automation and search.
- AI never makes religious judgments. No fatwa, no hadith grading.
- A person approves anything that matters before it reaches a user.
- When software suggests, the interface says so: suggested, draft, for review.
Quran and hadith text
Verses come from a verified text (Tanzil). They are copied, never typed or generated, and always shown with the surah name and ayah numbers.
Every hadith carries its collection and number. Where scholars differ on a grading, we show it with the scholar who gave it, never as our own judgment.
How we handle data
Your data is a trust. We never sell it, and we never use it to train shared AI models.
[CONFIRM] Where data is stored, who can see it, and how deletion works, for each product, before launch.
Working with us
[CONFIRM] We take on a small number of engineering and AI automation projects for organisations that serve students of knowledge, alongside our own work.
A good fit
- A problem where reliability is not optional.
- A team that wants to own what we build together.
- A scope that can start small.
How a project starts
- A conversation.
- A written scope.
- A fixed first milestone, with working software at the end of it.
Build something that has to hold.
Tell us what you are working on. We take on a small number of engineering projects alongside our own work, and we reply to every message.