We live in a state of cognitive fracture.
Your code lives in GitHub. Your research lives in Chrome. Your communication lives in Slack. Your thoughts live in Notion. These systems do not talk to each other. They are isolated silos of intelligence.
As we approach the AGI horizon, this fragmentation becomes a liability. AI models are generic geniuses they know everything about the world, but nothing about you. They lack the context of your specific history, your unique patterns, and your implicit knowledge.
Rverity is a self-hosted Knowledge Graph that unifies your digital existence. It is an infrastructure layer that sits below your applications, observing the data stream of your life. It vectors, links, and indexes every interaction into a single, navigable 3D graph.
It does not just "search" your history. It reconstructs your train of thought.
We adhere to a philosophy of sovereign computing. You run the stack. You own the weights. You control the graph.
- Runtime: Next.js 16 (App Router) on Node.js Edge.
- State Management: Server-side highly consistent state via TanStack Query.
- Visual Engine: React Three Fiber (WebGL) for high-performance 3D graph rendering.
- Vector Database: Supabase (PostgreSQL + pgvector) for high-dimensional semantic search.
- Caching: Upstash Redis for sub-millisecond context retrieval.
graph TD
subgraph Sources ["Data Ingestion (Extensions & Webhooks)"]
VS[VS Code Extension]
CH[Chrome Extension]
GH[GitHub App]
end
subgraph Core ["Rverity Core (Next.js Application)"]
API[API & Server Actions]
Embed[LLM Embedding Engine]
end
subgraph Storage ["State & Persistence"]
DB[(Supabase / pgvector)]
Cache[(Upstash Redis)]
end
subgraph Client ["Visual Interaction"]
UI[Dashboard Interface]
Graph[React Three Fiber 3D Graph]
end
VS -->|Source code, Editor context| API
CH -->|Browsing history, Highlights| API
GH -->|Commits, Issues, PR events| API
API <-->|Generate vector embeddings| Embed
API <-->|Semantic search & storage| DB
API <-->|Cache context| Cache
UI <-->|CRUD & Auth| API
Graph <-->|Query relationship nodes| API
- Ingestion: Passive observers in VS Code and Chrome capture text, code, and navigational intent.
- Synthesis: Data is chunked and embedded using OpenAI/Anthropic high-fidelity models.
- Graphing: Semantic relationships are established between disparate data points (e.g., linking a StackOverflow article to a specific Git commit).
- Recall: The Context Engine anticipates your needs, surfacing relevant memories before you explicitly query for them.
We assume you are comfortable with a terminal.
-
Clone the Protocol
git clone https://github.com/Ramrajnagar/Rverity.git cd Rverity -
Ignite the Engine
npm install npm run dev
-
Access the Interface Open
http://localhost:3000. The graph will begin assembling immediately.
We maximize for three variables:
- Speed: Latency is the enemy of thought. Every interaction must be instant.
- Privacy: Your thoughts are your own. We build for a local-first future where data never leaves your perimeter without explicit consent.
- Aesthetics: Tools for the mind should be beautiful. We reject the utilitarian drabness of enterprise software.
Refining the interface between human intent and machine execution.
- Neural Visualizations: Immersive WebGL graph navigation.
- Predictive Context: Agents that code with you, not just for you.
- Local Vaults: Fully offline, local-only vector storage.
- Brain-Computer Interface: Experimental typeless input streams.
Built with conviction by Ramraj and the Rverity Team.