GEOBEAT.xyz

Geographic decentralization is one of the most important but least understood properties of decentralized networks. Today, teams lack a consistent way to answer basic questions about where their infrastructure actually runs, what jurisdictions control it, and how dependent they are on a small number of cloud providers. The Geographic Decentralization Index (GDI) is a research-driven attempt to make this legible.

The index is built around three core dimensions:

• Physical Distribution — how widely nodes are spread across the surface of the world, measured through spatial entropy, effective region count, and clustering analysis.
• Jurisdictional Diversity — how exposed a network is to aligned legal systems or single-country control, quantified through entropy, effective jurisdictions, and legal-bloc correlations.
• Infrastructure Heterogeneity — how dependent the network is on specific clouds, ASNs, or datacenters, assessed through concentration metrics and provider diversity.

Design principles: open methodology, participatory governance, interpretable metrics, reproducible logic, defensible data sources, and an honesty about uncertainty rather than overconfidence. The index must be simple enough to understand, rigorous enough to trust, and flexible enough to adapt as the ecosystem evolves. Governance will be community-driven: all assumptions, weightings, and datasets are published openly, with an explicit process for critique, refinement, and versioning.

GEOBEAT is the interactive tool built on top of the index. It functions as a real-time geographic observatory for decentralized networks, enabling teams, researchers, auditors, and policymakers to explore where blockchain networks are actually concentrated. GEOBEAT visualizes the three sub-indices, provides diagnostic breakdowns (e.g., effective jurisdictions, cloud shares, clustering signatures), and exposes the underlying data and logic. It is designed to be a neutral measurement layer — something any L1/L2, node operator, or service provider can use to understand and improve their geographic decentralization profile.

The broader goal is to open up the field of network geography. In parallel to the index and dashboard, we are building a framework for multifactor location proofs — combining diverse signals, inference methods, and attestations — to improve the measurement and management of network geographies across the ecosystem.

We hope this work contributes to a more durable, open, user-centric web underpinned by blockchains and other decentralized networks.

We built Geobeat by stitching together a full research → pipeline → analysis → frontend loop in a very compressed timeframe. A large portion of the development process involved rapid iteration with Claude Code, ChatGPT, and v0 — using AI tools not as scaffolding, but as collaborative engineering environments. Design work happened in Figma and Replit’s design mode, which let us prototype UI interactions and animations while the backend pipelines were coming online.

For data ingestion, we used Armiarma from Miga Labs alongside Grafana and Prometheus, giving us a unified telemetry layer across Filecoin, Ethereum, and Polygon. This provided high-frequency node, provider, and network-geo metadata that we could normalize into a consistent schema.

The analytical layer is written in Python using GeoPandas, PySAL, SciPy, scikit-learn, and H3. This is where we compute spatial entropy, clustering metrics, effective region counts, jurisdictional diversity, and infrastructure concentration—essentially the core sub-indices of the Geographic Decentralization Index.

On the frontend, we used Next.js with Mapbox GL JS, D3, shadcn/ui, and Recharts. v0 and Claude Code helped us assemble the dashboard quickly, while we hand-tuned the core hero animation and data visualizations. We also took the Compounding Engineering plugin for a first spin to accelerate the more intricate UI flows—surprisingly effective for complex stateful components.

The hackiest parts were also the most rewarding: building a credible v0 of a scientific index while simultaneously prototyping the data pipelines, inference logic, and interaction patterns. The real victory wasn’t just the product—it was assembling a technical architecture that can realistically scale into a harmonized, research-grade index for geographic decentralization. We started with a broad research vision, then aggressively downscoped it into a functional MVP that still points toward the complexity of the full system.