Lead generation through tender parsing — with transparent scoring.

The manufacturer finds clients by analyzing tenders. Before us — manual monitoring of dozens of sources, missed key procurements, and late entry into projects when the pipe type is already locked in — and often it isn't their material.

Procurement and sales live in separate spreadsheets — there's no single view of the market.

The task — a repeatable stream of relevant leads and a unified view of the market.

• a.Daily monitoring of key sources.
• b.Project parsing before the public tender stage.
• c.Extraction of technical pipe attributes.
• d.Transparent scoring — why this is a lead, not noise.

Canonical registry consolidating dozens of tender sources into a single unified representation.

Product-fact extraction from procurement documents: product, material, diameter (e.g., PE100 SDR17 Ø160), length in linear meters, quantity.

Procurement card with metadata and business context: region, customer with tax ID, industry, pipe type, scope, work type, object type.

Score transparency instead of magic 90%. Explicit sub-criteria: starting price, selection method, deadline, pipe-type multiplier. The salesperson sees why a tender is relevant and can challenge it.

Summary description for each procurement — what's being purchased, relevance, customer requirements, documents.

Market meta-analytics. Who's winning, what volumes, segment-level revenue — for production and sales planning.

• a.Dozens of heterogeneous tender-document sources unified in one view.
• b.Extraction of technical product fields from unstructured text.
• c.Scoring with full decomposition: multipliers, reasons, weights — no black box.

• d.Market analytics layer on top of the collected data.
• e.Early entry — leads before the public tender stage.
• f.Parsing backbone reused in case 04 (pumps).