The data layer Web3 has been waiting for

Sam kamani

September 1, 2025

‍

When Chris Liu left venture capital to co-found ChainBase, he wasn’t just starting another data platform—he was taking on the messiest, most underappreciated challenge in Web3: organizing blockchain data at scale.

In Episode 176 of Web3 with Sam Kamani, Chris breaks down how ChainBase is tackling this head-on, building the connective tissue that helps developers, AI models, and applications make sense of the fragmented world of blockchain.

From VC to builder: following the data

Chris Liu's crypto journey started in 2016. He dabbled in Bitcoin, met early believers, and eventually backed blockchain startups as a VC. But over time, he grew frustrated with the recurring bottleneck: access to clean, usable data. So he built the solution himself—ChainBase.

Now, ChainBase supports over 200 blockchains and serves more than 20,000 developers. Its core promise? A unified, AI-ready data layer for Web3.

“Every blockchain is just a giant database,” Chris says. “But accessing it, indexing it, and actually using the data? That’s the hard part.”

Why on-chain data is so broken

The current state of blockchain data is like trying to build software on top of 200 different spreadsheet formats—each with its own quirks, labels, and inconsistencies.

ChainBase solves this by:

Aggregating data across chains
Standardizing formats
Structuring data specifically for machine learning
Offering real-time access via APIs and SDKs

This approach allows security firms, financial analysts, and DeFi platforms to build faster, smarter tools without reinventing the wheel.

Building for AI from day one

One of ChainBase’s standout features is that it doesn’t just serve data—it prepares it for AI. That means:

Labeled, structured datasets
Clean inputs for training models
Tools for AI-driven insights on-chain

Chris explains how their infrastructure enables a shift from “model-driven AI” to “data-driven AI” in crypto—a subtle but powerful transition. As he puts it: "The best models can’t fix bad data. If you want smarter AI in Web3, you need better inputs."

Real-world traction: who’s using ChainBase?

Security companies use ChainBase to flag suspicious transactions
Trading firms use it to generate signals across chains
AI tools rely on it to structure blockchain data for training models

One impressive stat: over 10,000 projects are currently powered by ChainBase’s infrastructure.

The killer app? It might already be here

While many projects chase the next breakout use case, Chris believes we already have one: stablecoins. Their role in seamless, low-fee global payments makes them the most obvious real-world value in blockchain today.

“Stablecoins are simple, useful, and everywhere,” he says. “That’s the definition of product-market fit.”

What’s next for ChainBase?

Chris teased a number of exciting updates:

A token generation event (TGE)
ChainBase’s mainnet launch
Open-source releases of crypto financial models
New tools for AI developers in Web3

But more than roadmap milestones, he emphasized the importance of building public goods: “We want to create infrastructure that everyone in the ecosystem can rely on.”

Final thoughts: a call to build

Chris closes with an invitation to developers and builders worldwide: contribute to the data ecosystem. Whether you’re launching a DeFi protocol, building AI tools, or just curious about blockchain data—ChainBase is open for collaboration.