From The Graph to Geo: How One Engineer Is Decentralizing Knowledge Itself

Remember when the web was supposed to be open and decentralized? Yeah, that didn’t exactly happen the way Tim Berners-Lee and the original architects envisioned it.

Instead, we got Facebook, Google, Twitter, and a handful of mega-platforms that hoarded our data, controlled the narrative, and turned the internet into a series of walled gardens. The value we collectively created got trapped inside proprietary databases, controlled by corporate gatekeepers.

Yaniv Tal has been thinking about this problem for nearly a decade. And after years building The Graph—one of web3’s foundational protocols—he realized something crucial: the infrastructure for decentralized applications wasn’t just missing pieces. It was fundamentally incomplete.

That realization led to Geo, a project so ambitious, so technically complex, and so philosophically grounded in the original vision of the web that it’s taken four years of heads-down development to get right.

The Problem We’re All Facing

Let’s be real: it’s nearly impossible to know what’s true anymore.

We’re drowning in information. Social media is chaos. Our systems for organizing and making sense of what’s happening in the world are broken. Misinformation spreads faster than corrections. Knowledge exists in silos, trapped behind different platforms, different companies, different incentive structures.

What if there was a better way? What if we could organize knowledge in a way that was transparent, verifiable, and collectively governed?

That’s the core insight behind Geo. It’s a network for organizing knowledge that we can actually trust.

Building the Impossible: Protocol Meets Consumer App

Here’s where things get really interesting—and really hard.

Yaniv describes the core challenge of Geo beautifully: it’s part web protocol and part consumer application. Those two things are fundamentally at odds with each other.

A web protocol’s job is to be perfect and never change. It’s the foundation. You can’t move it around once millions of people are building on top of it.

A consumer app’s job is to iterate constantly. Daily. Sometimes hourly. Based on user feedback, changing market conditions, and the relentless pursuit of a great user experience.

Doing both simultaneously? That’s the core tension Yaniv and his team have been wrestling with for four years.

Their approach is unconventional. Rather than building the protocol in isolation and then wrapping a UI around it, they took what Yaniv calls an “Apple approach”—thinking of the entire system as integrated, with every piece designed to work seamlessly with every other piece.

The Architecture: Pieces Coming Together

Geo isn’t a simple product. It’s a complex system with multiple interconnected layers:

The Knowledge Graph Standard. First, you need a way to represent information that’s more flexible and powerful than the semantic web’s aging RDF standard. Geo uses a graph data model that lets you structure free-form relationships between different things—more intuitive, more powerful.

The On-Chain Layer. This is where governance happens. Different participants can perform different actions: voting, signaling, upvoting, downvoting, approving, rejecting. It’s essentially a social layer operating through governance mechanisms.

Indexing and Database. Taking an immutable blockchain and making it queryable across history is genuinely difficult. Structuring this across space and time required serious engineering thinking.

The SDK. Developers need tools to build custom applications on top of Geo. That means building an entire development environment.

The Geobrowser. This is the consumer-facing application—a browser for this new verifiable web. Everything you see is attributable. You can trace the full provenance, see exactly who published what and when. Imagine Git for data, but for the entire web.

The Four-Year Approach: Learning in Public

Yaniv is a believer in getting things in front of users as quickly as possible. But he also understood something crucial: there’s a massive difference between getting user feedback and scaling something.

Once you have hundreds of thousands of users, making structural changes becomes nearly impossible. You’re locked in.

So they took a different approach. Real users were using Geo within six months. But before scaling, they gave themselves time to explore, to learn, to iterate.

They worked in roughly six-month cycles, applying learnings and making changes. The journey started with a native app built in React Native—a simple UI-only browser. They borrowed from what Notion was doing with block-based editors, creating something that felt familiar but was fundamentally different.

Then they iterated on the data layer. The first version used a relational model. Yaniv always suspected a graph data model was the right answer, but they needed to work through the details: type systems, ontologies, how rigid or flexible to make different things.

Rather than optimizing for a single use case, they deliberately sought out a wide cross-section: personal applications, business applications, community-based applications. This diversity of use cases forced them to build something genuinely general-purpose.

The Bigger Vision

What Yaniv is building isn’t just a product. It’s a realization of what the original web architects actually wanted the web to be.

The semantic web vision was radical: knowledge and information should live outside the context of any single application. The web itself should be an open, interconnected graph. Not a web of pages, but a web of data.

That vision got derailed. The web re-centralized around a handful of giant tech platforms. Ethereum and blockchain protocols are trying to reverse that—to re-architect our core coordination systems so they’re open, transparent, interoperable, and aligned with those original intentions.

Geo is the next step in that evolution. It’s taking the lessons learned from building The Graph, combining them with a genuine understanding of what users actually need, and creating infrastructure for a more trustworthy, decentralized information ecosystem.

Why This Matters

We’re at an inflection point. The current systems for organizing and sharing information are breaking down. Misinformation is rampant. Trust is eroding. The platforms that control our information are facing increasing scrutiny and regulation.

There’s a real opportunity to build something better. Something that’s collectively governed. Something where you can verify the source of information. Something where the value created belongs to the community, not a corporation.

Yaniv and the Geo team have spent four years making that vision real. And according to Yaniv, we’re now at “a beautiful point of convergence where the pieces are really snapping together.”

That’s worth paying attention to.

Want to hear the full story? Listen to the complete episode with Yaniv Tal on Code Story to dive deeper into the technical decisions, the governance challenges, hiring for a decentralized organization, and what’s next for Geo. It’s a masterclass in building infrastructure for a better web.