Tactical intelligence for Bitcoin

Intelligence that learns how Bitcoin behaves.

TrackBTC is a tactical intelligence platform trained by AI agents. Swarms of agents discover, test and refine patterns in how Bitcoin is held and moved.

Put that knowledge to work for your team and your own AI systems to recognize routines, anomalies and meaningful change.

Discuss enterprise intelligence See what behavior reveals

Already a subscriber? Open current Intelligence

Measured finding Binance cluster · 21 Sep – 27 Sep 2026 Binance cluster keeps 5,623.14 BTC unspent while its balance falls by 24,828.51 BTC.

Balance reduction · 21–27 September

24,828.51 BTC

The tracked funds received on 15 September remain entirely unspent:

Received on 15 Sep 2026
5,623.14 BTC
Unspent at 00:00 UTC, 28 Sep 2026
5,623.14 BTC

This cluster receives no funds from other clusters during 21–27 September. It sends 24,828.51 BTC while the tracked 5,623.14 BTC received on 15 September remains unspent.

Same week, different behavior

How Binance, Coinbase and Bitfinex clusters use their funds differently.

In the week of 21–27 September 2026, the Binance cluster keeps earlier funds unspent while reducing its balance. The Coinbase cluster handles growing transfer volumes with a similar end-of-week balance. The Bitfinex cluster adds funds for a third consecutive week.

Each example covers one cluster: a group of Bitcoin addresses linked by observed spending and attributed to the named exchange.

Flow and balanceCoinbase cluster · 21 Sep – 27 Sep 2026

Coinbase cluster sends 87,615.83 BTC in a week and ends with 351.85 BTC.

Received and sent by complete week
Week ofReceivedSent
07 Sep51,219.87 BTC51,209.60 BTC
14 Sep67,705.00 BTC67,688.52 BTC
21 Sep87,609.60 BTC87,615.83 BTC

The cluster receives 87,609.60 BTC during the week. Its balance falls from 358.35 to 351.85 BTC, while incoming and outgoing volumes both rise.

Why it matters. The similar starting and ending balances conceal the scale of transfers handled during the week.

Read the finding

AccumulationBitfinex cluster · 21 Sep – 27 Sep 2026

Bitfinex cluster increases its balance for three consecutive weeks.

Net change after sending and fees
  1. Week of 07 Sep226.72 BTC
  2. Week of 14 Sep2,065.17 BTC
  3. Week of 21 Sep2,086.86 BTC

In the latest week, the balance rises from 17,787.43 BTC to 19,874.28 BTC. The cluster receives 2,086.86 BTC and sends 0.00 BTC to other groups.

Why it matters. The balance grows in each week shown, and no funds leave for other groups during the latest week.

Read the finding

Trained by AI, for AI

AI agents test new patterns against recorded Bitcoin activity.

Learning loopHow agents discover and improve patterns
  1. 01DiscoverAgents surface candidate patterns in how clusters behave.
  2. 02TestEach candidate is checked against what actually happened on-chain.
  3. 03RefineNew evidence sharpens, expands or revises what the platform recognizes.
  4. 04RetainTested patterns become available for other agents to use and improve.

Retained knowledge feeds the next round of discovery.

AI agents examine Bitcoin behavior from different angles, test what they find and refine what the platform knows. Each round builds on the knowledge already learned.

Findings for your analysts and AI systems

Your AI systems
Every finding is stated as exact figures with explicit dates and a defined scope. An agent can reason with it without reinterpreting raw transactions.
Your analysts
The same finding has a permanent page your team can open, cite and challenge.
Our own agents
Each retained pattern becomes context for the next round of learning.

What the platform learns to recognize

Recognize recurring behavior and the changes that break it.

The platform compares transfer sizes, timing, balances and counterparties with each cluster’s history. It identifies routines that continue and measures where activity changes.

What is routine for one exchange is not the template for another.

01Recurring schedules
When does this group usually move funds, and which expected transfers did not occur?
02Balance retention
Which incoming funds remain unspent while other funds leave?
03Reserves
How much does a cluster usually hold, and how does that compare with its transfer volume?
04Accumulation
Does the balance grow for several weeks, or only after one large incoming transfer?
05Counterparties
Who sends to it, who receives from it, and who has stopped?
06New routes
Which destinations and sources appear for the first time?
07Operational changes
Did transaction counts, sizes or timing shift in a way that signals a new process?
08Breaks in pattern
When do familiar transfer sizes, schedules or destinations change?
Public · daily

Latest daily alerts

Large transfers by named and unnamed clusters, with their daily totals and balances.

A cluster is a group of Bitcoin addresses linked by observed spending.

See all daily alerts

Public · latest completed week

Weekly Intelligence

Compare changes in balances, transfers and counterparties with earlier activity. New findings are published after each completed week is verified.

Daily alerts
Public as each alert is published
Weekly Intelligence
Public · latest completed week
Monthly profiles
In the app · completed-month context
Explore the app Read the public edition

Put it to work

Bring behavioral intelligence into your operation.

  1. Current Intelligence

    Read the latest completed week’s findings alongside monthly cluster profiles.

    Open current Intelligence
  2. Enterprise intelligence

    Agree coverage of the clusters, counterparties and behaviors your team and your AI systems rely on.

    Discuss enterprise intelligence
  3. Six-week pilot

    Evaluate one Bitcoin workflow and finish with a record of findings, checks and decisions.

    Explore a six-week pilot

Go deeper: worked examples in our articles, research, daily alerts and the public Intelligence edition.

CWOS is our first-party decision platform, built on this intelligence. Explore CWOS