Zakupki Keyword Layer — Neo4j Graph Analysis
Date: 2026-04-18
Graph layer added: :KeywordTenderWinner (8,237 nodes) + :SOUGHT_KEYWORD_TENDER edges (1,306)
Source: pipeline/data/zakupki_unified_inns.json (dedup of /tmp + pipeline jsonl keyword sweeps)
Ingestion: pipeline/neo4j_ingest/ingest_zakupki_keywords.py
Queries: pipeline/zakupki_graph_queries.py
Results: pipeline/data/zakupki_graph_analysis.json
Headline
- 8,237 zakupki keyword-winner INNs are now first-class graph nodes (previously JSON-only)
- 1,306 INN matches to existing
Organizationnodes — 294 MORE matches than the prior JSON-only join against target_scorecard (1,012). The entity graph is a wider net than the scorecard. - 148 already-sanctioned keyword winners (vs 143 from JSON join — more comprehensive sanctioned flag propagation via graph edges)
- 40 designation candidates surfaced at 2-hop SDN proximity, with the top 10 each reaching 600-1,114 distinct sanctioned nodes through procurement/litigation edges
- 6,931 unmatched INNs (orgs winning keyword tenders but not yet in our Russian entity graph) — this is the expansion universe
🔒 Premium brief — request full access
This brief's evidentiary detail, specific targets, and designation recommendations are available to subscribing compliance teams, law firms, and government analysts. The teaser above shows the headline methodology; the full document includes case-number citations, target INNs, SDN-adjacency numbers, and recommended secondary-designation scope.
Email derek.linz@linzalytics.com or click below to request access. Include your organization, role, and intended use.
Request access → See pricing tiers