The integrity record
Our data is clean — and we self-heal, honestly
Athath is a discovery layer over a real catalog we gather from hundreds of Egyptian stores. That means data does break sometimes: an image dies, a product is pulled from a site, a price comes out wrong. This page explains honestly how we detect that, correct what we can — and says with equal plainness what we can't.
How we keep the data clean
We detect what broke
A recurring check hunts for dead images (links that 404 or aren't an image), products that vanished from their source, and values that look obviously wrong to a human (a financing price, a logo image instead of the product, an empty title).
We quarantine before it shows
Data that fails the plausibility test (an impossible price, junk dimensions, an empty title) is held out of search and ranking — a deterministic, unit-tested gate — before anyone can ever see it.
We correct what we can
When a source changes or a product returns, a re-crawl picks up the correct value and updates the record. The healing is gradual — it covers the catalog lap by lap, not all at once.
We state the gap plainly
When an image has died and isn't fixed yet, we show a clean fallback and say it's unavailable right now — we don't hide the product to look perfect. A product pulled from its source is marked "no longer available" rather than left as a phantom.
Live signal (aggregate)
These figures are aggregate, catalog-wide only — never any single source's internals — and carry their as-of date. If the live signal is unavailable right now, we say so rather than invent a clean number.
As of: 7/21/2026, 3:19:41 PM
What we can't fix (plainly)
- If a store's own site is down or has pulled a page, we can't restore its data — we mark it unavailable rather than guess.
- Dimensions or details a store never published, we show as "not provided" — we don't fabricate them as fact.
- Healing takes time: the catalog is large and the re-crawl is gradual — so you may see something stale for a while before it updates.
Go deeper