Diamond demand matching, step by step

1. A request arrives as plain text

In a monitored Telegram trading group a buyer posts something like “looking for 1.20–1.50 round, F–H, VS, GIA” — or the same thing in Hebrew, in a different order, with abbreviations. BrilliantBot reads both Hebrew and English, because that is what the groups it monitors are written in.

2. The request is parsed into criteria

The free text is turned into structured fields — shape, weight range, colour range, clarity range, and whatever else the buyer specified. A request that is too vague to parse stays vague; the bot does not invent a specification the buyer did not write.

3. Criteria are compared to uploaded inventory

Matching only works on stock the platform knows about, so the first step for a new dealer is uploading a stock list. Each parsed request is then compared against those stones. The comparison is on the attributes the buyer named, which is why a clean, current stock list produces noticeably better matches than a stale one.

4. Relevant users get notified

When a request lines up with stones you hold, you get a notification in Telegram with the request and the matching stones, so you can answer the buyer while the request is still fresh. No match means no message — the point is to reduce what you have to read, not add to it.

Honest limits

• Coverage is limited to groups the bot has been added to. There is no all-of-Telegram view.
• A match is a signal of interest, not a sale. Price, terms and trust are still yours to negotiate.
• Parsing free text is imperfect; ambiguous requests can be matched loosely or missed.
• Out-of-date inventory produces matches on stones you no longer have.

Cost and next step

A new account gets a full 14-day trial, and after that it is $150 per month. If you want the wider market context first, read B2B diamond trading on Telegram, the FAQ, or see who builds BrilliantBot.