User Guide

Welcome to Hōkū

Hōkū is the first real-time, AI-powered trading intelligence platform built entirely for on-chain markets. It transforms the raw flow of blockchain transactions into clear, actionable insights, helping traders understand what’s happening, why it’s happening, and where the next opportunity might be forming.

Unlike traditional explorers or dashboards, Hōkū doesn’t just show data — it interprets it. Every feature in the app is powered by proprietary machine learning systems developed by Deep3 Labs, trained on hundreds of millions of transactions to model the behavior of on-chain traders themselves. The result is a new kind of trading interface: fast, intelligent, and deeply aware of how money actually moves across networks.

Hōkū is designed for everyone from individual traders to advanced analysts. Whether you’re tracking your own portfolio, studying the behavior of profitable wallets, or monitoring entire market segments in real time, this guide will help you understand each part of the platform and how to make the most of it.

How Hōkū works

At its core, Hōkū runs on a proprietary data architecture designed by Deep3 Labs to process and analyze blockchain data faster than traditional explorers. All training data is streamed directly from Deep3 Labs’ own network nodes, ensuring near-zero latency and complete transparency.

Machine learning models continuously classify and embed every active wallet into behavioral “clusters,” mapping the relationships between traders in a multidimensional space. These same embeddings are then used across modules: to power the 3D Wallet Cloud explorer, to identify cluster-level token trends, and to feed the recommender systems behind personalized token relevance and trade signal alerts.

Every prediction, trend, and visualization in Hōkū is ultimately connected to that shared foundation providing a live, learning model of the blockchain’s traders themselves.

How this guide is organized

Each module in this user guide follows a consistent structure to make it easy to learn how every part of Hōkū works, from quick overviews to deeper technical explanations.

• Module Summary Each section begins with a high-level overview explaining what the module does, how it fits into the broader Hōkū ecosystem, and what types of insights it provides. The summaries are designed to help you understand why the feature exists and how it can be useful before diving into the details.

• Features This section breaks down every control, button, or data display within the module. Each feature is described clearly, with screenshots where necessary, so you know exactly what information you’re looking at and how to interact with it.

• Usage Tips These are real-world examples that show how traders and analysts can apply the data or tools in practice. Rather than just describing what a feature does, this section focuses on how to use it strategically to uncover opportunities, validate ideas, or react to changing market conditions.

• FAQ A collection of the most common questions and clarifications about the module, including how to interpret specific metrics, where the data comes from, and how updates occur. Each FAQ entry is short, direct, and designed to address confusion before it arises.

• How It Works (select modules only) For readers interested in the engineering or data science behind Hōkū, some modules include a deeper technical dive. These “How It Works” sections explain the underlying architecture and algorithms in accessible language, combining technical accuracy with plain-language explanation.

Together, these sections provide both a user’s guide and a technical reference, helping you get value from Hōkū whether you’re exploring casually, trading actively, or studying its machine learning systems in detail.

Support

Throughout the app, you'll see a circled question mark icon that launches context-specific help docs directly in the interface. If you get stuck, this is the best first step.

But, for any other questions, or just to share feedback, reach out to us via one of our community platforms or send us an email at [email protected].