We weave the
chaotic into
meaning.
Mastum is a digital loom — we take raw data, handwritten pages and tangled workflows and weave them into AI that's structured, cohesive and native. Our first threads: Edubbly and Tamkar.
Mastum means “loom.”
Long before software, the loom was the machine that turned chaos into cloth — a frame disciplined enough to hold loose, independent threads and weave them into something whole and strong. We took its name on purpose.
Mastum is that loom, rebuilt for intelligence. We gather the raw and the scattered — unstructured data, handwritten pages, tangled business processes — and weave them into AI ecosystems that are structured, cohesive and natively intelligent. Edubbly weaves a photographed page into a reasoned grade; Tamkar weaves product photos and customer messages into a storefront that runs itself.
We don't just build apps. We're the engine that structures the chaotic — weaving intelligence, thread by thread.
We build the layer between language and machines.
𒈠𒊏𒁕𒌝 — what we makeEdubbly
Photograph a handwritten paper and Edubbly reads it — Arabic, English and handwritten mathematics alike. It grades by meaning, then returns a PDF that lays out every correction with the reasoning behind it.
From a photo to a reasoned PDF.
𒁾𒌓 — how Edubbly worksout of 100
Snap the paper as it is.
No scanner, no template. A phone photo of the desk — tilted, shadowed, pencil-smudged — is enough. Edubbly reads Arabic script, English prose and handwritten working straight off the page.
Graded by meaning, not keywords.
The engine weighs what the answer actually says — ideas, method, intent. A right thought in the wrong words still earns its marks; a memorized keyword in the wrong place doesn't.
A report that argues its case.
Minutes later a clean PDF comes back: every correction, the mark it cost and the reason why — ready to hand to the student, the parent or the ministry.
Move your cursor across the text.Every correction, explained.
At a glance, a graded paper just looks marked. Move Edubbly's light across it and the reasoning behind each correction surfacesEdubbly surfaces the reasoning behind each correction — the same notes it writes into the PDF.
The committee has reviewed each proposal, and it believe the impact in students are significant and well documented.
Tamkar
Tamkar runs your Instagram shop. Build a product database by simply photographing what you sell — AI reads each image into a clean catalog — then a bot greets customers in a warm Iraqi dialect, quotes prices and takes orders, so no DM goes unanswered.