Coinator answers three questions: Who owns these addresses? Where did the money go? How risky is the counterparty? Every answer shows what it is based on and how certain it is. More than 40 documented methods, updated daily.
Transparentnost
Proč transparentnost?
Na rozdíl od mnoha jiných nástrojů pro sledování transakcí Coinator rozhodně sází na transparentnost místo tajemnosti. Z našich algoritmů, heuristik a analytických postupů neděláme žádné tajemství - naopak: všechny použité metody jsou otevřeně zdokumentované a kdykoli k nahlédnutí.
Více než 40 jasně sledovatelných postupů - mimo jiné multi-input heuristiky, analýza návratových adres, detektory CoinJoin a PayJoin a četné filtry - zajišťuje maximální sledovatelnost, vysokou spolehlivost a soudně přípustné výsledky na evropské úrovni ochrany osobních údajů.
Five building blocks
Methods at a glance
Five building blocks that together carry the case: grouping addresses, making evidence quality visible, understanding transactions, following trails across the chain boundary, naming and assessing actors.
01
🔗
Grouping addresses
Which addresses belong to the same person or the same service? Coinator combines proven blockchain-analysis methods and places every attribution in one of three confidence tiers. You decide how conservative the analysis should be.
Highest tier only with evidence from the blockchain itself
Extended tiers only when several independent indications agree
Protection against misattribution through obfuscation transactions (CoinJoin, PayJoin)
Every analysis reproducible via link, even months later
02
🧾
Evidence quality made visible
Not every link is equally robust. Coinator shows for every edge in the graph where its confidence comes from. You see at a glance what will hold up in court and what remains a lead.
Evidence from the blockchain itself, across all transactions since 2009
Statistically tested evidence, calibrated against independent reference data
Your own decision, documented in the case file
Origin of batch payouts resolved back to the payer
03
🏷️
Understanding transactions
Behind every transaction there is a pattern: an exchange payout, a mixer, a Lightning channel, an inscription. Coinator recognises these patterns across the entire blockchain and shows them where they matter in an investigation.
Obfuscation transactions detected and flagged
Exchange payouts, deposits and mining proceeds attributed
Lightning Network and Ordinals visible instead of a blind spot
Updated nightly, no stale data
04
🌉
Across the chain boundary
Anyone who wants to cover their tracks switches blockchains. Coinator follows swaps through the major cross-chain protocols and links deposit and payout, each with a stated confidence.
THORChain, Maya, Chainflip and further bridges covered
Source and destination chain shown for every transaction
Older swaps captured retroactively as well
Filters in search and graph
05
🚦
Who is it, and how risky
Names and risk in one rule set. The entity register tells you who is behind an address. The risk indicator turns that into an assessment you can trace back to the payment contact that triggered it.
More than 1,000 services with category, jurisdiction, KYC and sanctions status (OFAC, EU, UK)
Sanctions are time-aware: a payment before the listing does not count as a breach
Internet reports on addresses shown as findings, separate from the assessment
Similarity search finds further accounts of the same operator
Machine learning
Calibrated evidence instead of a black box
Machine learning complements the methods where rules alone cannot decide. Every model delivers a tested probability that remains visible as evidence.
01
Tested models instead of gut feeling:
Where rules alone cannot decide, a statistical model that has been tested against independent reference data makes the assessment. The result enters the graph as evidence with visible confidence, never as an anonymous number.
02
Majority instead of a single opinion:
Instead of trusting a single rule, several independent indications vote. Only when they agree is an attribution considered established.
03
Cluster profiles and similarity search:
Every large cluster receives a profile derived from its behaviour. The similarity search finds clusters with a comparable profile, for example further accounts of the same operator.
04
Research with the BCCM institute:
Together with the Blockchain Competence Center Mittweida (BCCM), an institute of Mittweida University of Applied Sciences, we develop and trial the latest analysis methods, including AI-based ones. Research results enter Coinator only after successful trials. What runs inside Coinator is documented and visible in the application.
Otázky k metodice?
Rádi vám vysvětlíme, které postupy se nejlépe hodí pro daný případ.