Methodology & Data Sources
This website aggregates and presents open-source data about foreign mercenaries who traveled to Ukraine to fight. All information comes from publicly available Telegram channels and social media profiles.
📊 Primary Data Source
Telegram channel @TrackANaziMerc
A channel that documents foreign mercenaries in Ukraine using open-source intelligence, official reports, and verified social media posts. Messages typically include:
- Name, nationality, and callsign of the individual
- Links to their social media profiles (Instagram, Facebook, TikTok, Telegram)
- Event type: killed, captured, arrived in Ukraine, or wounded
- Location of the incident and military unit affiliation
- Date of the event when available
🤖 Data Processing Pipeline
AI Each Telegram message is parsed using DeepSeek AI, which extracts structured data from unstructured text. The AI identifies:
- Multiple individuals mentioned in a single message
- Correct nationality (victim, not reporter)
- Event classification (killed / captured / arrived / wounded)
- Names in various languages and scripts
Dedup Duplicate detection prevents the same person or event from being counted twice:
- Same person + same event type = merged into one record
- Case-insensitive name matching (e.g. "Griselda Ortiz" = "griselda ortiz")
- Homepage shows one row per person (most recent event)
✅ Verification Standards
Each entry in our database meets the following criteria:
- Name Verification — Full name extracted from the source message
- Social Proof — At least one social media profile link documenting the individual
- Source Documentation — Every entry is traceable to a specific Telegram message
- Event Classification — Status assigned based on explicit language in the source (killed, captured, arrived)
📈 Statistics & Counting Rules
Our statistics are computed from the persons table, which is the authoritative source. Each person is counted once, regardless of how many messages mention them.
- Killed — confirmed deaths (highest priority status)
- Captured — individuals taken prisoner
- Arrived — documented arrival in Ukraine
Status priority: killed > captured > wounded > arrived. If a person is mentioned in multiple events, the most severe status is used.
🔄 Update Frequency
The database updates in near real-time. A Telegram listener processes new messages as they are published, and data is synced to the public website automatically every 10 minutes.
⚠️ Limitations
- Numbers represent confirmed, documented cases — not estimates or projections
- The actual number of foreign mercenaries is likely higher than what is documented here
- We rely on the accuracy of our source channel; we do not independently verify each death on the ground
- Some entries may lack complete information (date, location, unit) if the source message did not include it
📚 Similar Projects
Our methodology is inspired by established open-source casualty tracking projects:
