Methodology

1. Source scope

The only demand source is the diamond trading groups the bot has been added to. There is no crawling of the wider internet, no scraping of third-party marketplaces, and no access to groups the bot has not been invited into.

This is a real and permanent boundary, and it is the reason no full-market, industry-wide or "all of Telegram" coverage is claimed anywhere. What the platform observes is a defined subset of trade conversation.

2. Identifying a request

Most traffic in a dealer group is not demand. Offers, photographs, price discussion and ordinary conversation all flow through the same channel. Each message is therefore assessed for whether it expresses a buyer looking to purchase. Messages that do not are discarded from the matching pipeline.

3. Parsing and criteria normalisation

A request identified as demand is parsed into structured fields — typically shape, weight range, colour range, clarity floor, certificate preference and quantity — from text that may be Hebrew or English, abbreviated, and inconsistently formatted.

Normalisation maps varied notations onto one vocabulary so comparison is possible: differing weight formats, lettered grade ranges written several ways, and laboratory names in short form.

One rule governs this stage: nothing is inferred. If the buyer did not state clarity, no clarity requirement is recorded. A vague request remains vague, which widens the match set rather than narrowing it on a guess.

4. Matching against dealer inventory

Normalised criteria are compared against the stock lists dealers have uploaded. Because most criteria are ranges, this is interval logic: a stone qualifies when its weight falls inside the requested band, its colour and clarity meet the stated floors, and its certificate is acceptable to the buyer. Criteria the buyer did not state are not tested.

Only stones present on an uploaded list can be matched. Inventory the platform does not know about is invisible to the process.

5. Notification

Each dealer with a fitting stone receives a notification containing the request and their matching stones, delivered through the Telegram bot. Dealers without a fit are not notified. BrilliantBot can also support buyer notification and direct-message flows, but it does not autonomously quote, negotiate commercial terms, accept a deal, hold funds or settle payment on a dealer's behalf.

Known failure modes

• Out-of-scope demand. Requests posted in groups the bot is not in are never seen.
• Ambiguous requests. Requests that are unclear, contradictory or written in unusual shorthand can be matched too loosely or missed altogether.
• Classification error. An offer can be misread as a request, or a genuine request skipped.
• Stale inventory. Stones sold, placed on memo or repriced without the stock list being updated will still match, wasting the dealer's time.
• Data quality on upload. Incorrect grades or missing attributes on a stock list propagate straight into wrong or absent matches.
• Timing. A buyer may have already covered their need before a dealer responds.

What a match does not mean

A match is not a sale, an order, an offer or a commitment by either side. BrilliantBot is not a marketplace, exchange, broker, escrow service or payment processor: it does not set price, negotiate, hold funds or settle transactions, and it guarantees no commercial outcome. It also claims no endorsement by any trade organisation.

Read next

• Diamond demand intelligence — the category this method serves.
• Diamond inventory matching — step 4 in more depth.
• Glossary — every term used above, defined.
• Industry references — independently published sources concerning BrilliantBot.
• About — who builds the platform and independent references.