Notes from the workbench
What we are learning while building Madeenan
Field notes on Islamic AI, source trails, careful answers, and the small product decisions that decide whether a source-backed answer feels trustworthy.
July 2026 · 5 min read
Islamic AI safety checklist
A practical checklist for teams building Islamic AI features with citations, source review, cautious copy, and human escalation.
How to add cited Islamic answers to an app
A practical product guide for adding Islamic AI answers with visible source trails, careful copy, and reviewable citations.
April 2026 · 5 min read
Madeenan vs raw Quran APIs
A practical comparison for teams deciding between raw Quran APIs and a source-backed Islamic AI layer with citations.
March 2026 · 4 min read
Why Islamic AI needs source citations
A note on what breaks when fluent models answer religious questions without a source trail, and why Madeenan treats citations as the product surface.
June 2026 · 4 min read
Picking the right LLM for your Islamic app
A practical note on choosing model behavior when Islamic answers need restraint, citations, and source review.
May 2026 · 4 min read
The hard part of Islamic AI is not the model
A February field note from building Madeenan: the model matters, but source trails and restraint decide whether an answer deserves trust.
February 2026 · 6 min read