AIAny
AI Infra2026
Icon for item

WorldWideView

Visualizes live global data on a CesiumJS 3D globe via a plugin-driven pipeline and real-time WebSocket DataBus. Supports dynamic plugin marketplace, an opt-in Agent Bus for external LLM/MCP control, and self-hosting with Docker and PostgreSQL.

Introduction

Most geospatial viewers are either static map layers or tightly coupled data pipelines; WorldWideView flips that tradeoff by decoupling acquisition from rendering so independent data sources can stream in as loadable plugins while the globe remains responsive. That design makes it practical to run many high-frequency feeds (aircraft, maritime, event data) and let external controllers — including LLM-backed MCPs — operate the UI without touching core rendering code.

What Sets It Apart
  • All-Bundle plugin architecture: plugins are dynamically imported and run independently of the core viewer, so new data types can be added without rebuilding the frontend. This means operators can onboard novel feeds quickly and isolate faults to a plugin.
  • Real-time DataBus + chunked rendering: a typed WebSocket DataBus feeds a memoized Zustand store and chunked primitives to Cesium, enabling high-frequency updates with horizon culling and 3D stacking to keep frame rates usable on large entity counts.
  • Agent Bus (opt-in): an HTTP+SSE control surface intended for MCP/LLM servers lets external agents fly the globe, toggle layers, and select entities. That makes WorldWideView suitable as a visualization/actuation client in agent-driven workflows while remaining opt-in for security.
Who It's For & Tradeoffs

Great fit if you need a self-hosted, extensible situational-awareness frontend that ingests many live feeds and you want a clear separation between data seeders and the renderer. Look elsewhere if you require a turnkey commercial SLA, proprietary satellite imagery licensing, or a minimal 2D map UI — WWV is optimized for rich 3D visualization and developer extensibility, not for zero-config SaaS simplicity.

Where It Fits

WorldWideView sits between data-engine backends (pollers, streamers, MCP servers) and end-user visual workflows: it is the client-facing, plugin-driven visualization layer that can be integrated into incident rooms, analyst tooling, or agent-driven automation stacks. Paired with the community wwv-mcp reference, it becomes a programmable globe for LLM agents.

How It Works (brief)

The repo provides a Next.js frontend (App Router) that consumes a typed DataBus over WebSocket, hydrates a Zustand store, and renders primitives via CesiumJS. The platform encourages a monorepo plugin model (wwv-plugin-sdk) and ships deployment guidance for Docker + PostgreSQL. The Agent Bus is disabled by default and requires explicit configuration when used.

Information

  • Websitegithub.com
  • Authorssilvertakana
  • Published date2026/03/04

Categories

More Items

GitHub
AI Infra2025

Defines a vendor-neutral JSON/YAML semantic model specification and tooling to exchange metrics, dimensions, lineage and other business semantics across analytics, AI and BI platforms; includes a core spec, validators, converters (dbt, GoodData, Salesforce) and example models.

GitHub
AI Train2025

An asynchronous, high-throughput framework for large-scale reinforcement learning and agentic training that scales to 1T+ MoE models and 1000+ GPUs, with native verifiers integration, end-to-end SFT/RL/evals, and Slurm/Kubernetes deployment; requires NVIDIA GPUs.

GitHub

Runs a self-hosted meeting bot and transcription API that joins Google Meet, Teams and Zoom and streams speaker-attributed transcripts in real time. Compiles meetings into a git-backed Markdown workspace and runs sandboxed agents on your infrastructure; Apache-2.0 and air-gap capable.