Crypto Wallet Intelligence: Tracing, Analysis, and Compliance

Wallet intelligence means reading everything an address has done on a public ledger, then judging what kind of user is behind it and how much risk it carries. The raw material is simple: every transfer in, every transfer out, the tokens, the dates and the other side of each payment.

What it is used for

• Compliance checks. An exchange looks at where a new customer's funds came from before it takes the deposit.
• Audits. A finance team checks today's balance against the full history of what came in and went out.
• Mapping a group. Once one address is tied to a scam, the addresses it deals with most are the next ones to look at.
• Spotting odd moves. A payment far larger than usual, or one sent to a mixer, stands out against the past pattern.

A worked profile

Here is one Ethereum address, 0x41Bc7d0687e6Cea57Fa26da78379DfDC5627C56d, profiled from Bitquery's Ethereum index on 2 October 2026. We do not know who owns it.

Measure	Result
First transfer in	9 June 2019
First transfer out	21 June 2020
Transfers in	13,643, from 1,655 senders
Transfers out	9,786, to 1,160 receivers
Different tokens received	1,431
Different tokens sent	959
Last transfer in	1 October 2026

Three things can be read from the history.

It is a DeFi user. On 1 January 2024 the address received about 795 swETH in three transfers. On the same day it also received tokens whose names mark them as yield products built on swETH. That is the footprint of someone who stakes and farms yield.

Its large payments out came in a burst. The table shows the ETH it sent in the first half of 2024.

Month in 2024	ETH sent	Largest single transfer
January	635.33	150
February	319.00	300
March	633.00	500
April	349.36	200
May	1.00	1
June	9.35	6

A payment of 500 ETH would look alarming on its own. Set against four months of large transfers it fits the pattern, and the quiet months after it are the change worth a second look.

The token count overstates what it chose to hold. Anyone can send a token to any address, and busy addresses collect many they never asked for. A count of 1,431 tokens says more about how visible the address is than about its owner's taste.

An earlier version of this post said this address sent out only ETH, and never more than 37.7 ETH at once. The full history shows otherwise.

How to build a profile yourself

The quickest route is the Bitquery MCP server. You connect it to Claude, ChatGPT or Cursor and ask in plain English.

Prompts that cover the four uses above:

• "What role does this address play?" The tool behind it, eth_address_flow_summary, sorts an address into consolidator, distributor, relay or hub.
• "Is this address labeled as an exchange, a mixer or a scam?" This reads Bitquery's address labels.
• "Where did the funds from this address go? Stop at the first exchange."
• "What does this trader hold, and what is their profit and loss?"

Tracing and labels sit on the AI Investigation plan at $149 per month, and a trial comes with 100 free credits.

Developers can pull the same history through the V2 transfers API.

Red flags to look for

• A transfer many times larger than anything in the address's past.
• Funds sent to a mixer, or received from one.
• Many fresh addresses that each pay in once and never appear again.
• Money that comes in and leaves within minutes, again and again.

None of these proves wrongdoing. Each one is a reason to look closer before you act.

Related reading

• How to audit blockchain wallets
• Cryptocurrency money laundering explained
• Tracing Bitcoin transactions and addresses

First published in July 2024. Updated on 2 October 2026 by the Bitquery team: the example was worked out again from Bitquery's Ethereum index, and AI tools helped run the queries and redraft the text. Bitquery sells the MCP server and the address labels named above.