Knowledge Graphs
Structured concept relationships with thesaurus-driven automata. Semantic matching that understands your domain.
Terraphim AI
Semantic search across your knowledge. Private by design. Powered by knowledge graphs.
Project statistics
29
Integrated Crates
8+
Data Source Types
5
Relevance Algorithms
<2s
VM Boot Time
Built for Knowledge Workers
Connect, search, and understand your data across every source you use.
Haystack Integrations
Confluence, Jira, Discourse, JMAP email, local files, and more. Each becomes a searchable node in your graph.
Role-Based Profiles
Define distinct search contexts with custom relevance functions, knowledge domains, and source configurations.
AI Integration
Connect Ollama for local models or OpenRouter for cloud inference. Summarisation, chat, and intelligent descriptions.
Firecracker VMs
Sub-2 second microVM boot for secure command execution. Sandbox untrusted operations without overhead.
MCP Protocol
Full Model Context Protocol server. Expose autocomplete, text processing, and graph connectivity as AI tools.
Privacy Is Not a Feature
Your knowledge graph, your search index, your queries -- everything runs on your device. Terraphim does not phone home. There is no analytics endpoint, no usage tracking, no cloud dependency. Privacy is the architecture.
-- Terraphim Design Principles
Terraphim is built with Rust for performance and safety. The persistence layer supports memory, DashMap, SQLite, and S3 backends -- all configurable, all local-first. Cache warm-up happens transparently between layers.
Optional external connections (Confluence, Jira, email) are explicit, user-initiated, and authenticated with your own credentials stored in 1Password.
Quickstart
Terminal
git clone https://github.com/terraphim/terraphim-ai
cd terraphim-ai
cargo build --release
# Start the server
cargo run --release -- --config terraphim_engineer_config.json
# Or launch the TUI
./target/release/terraphim-agent
Desktop Application
cd desktop
yarn install
yarn run tauri dev
API Usage
curl -X POST http://localhost:8080/documents/search \
-H "Content-Type: application/json" \
-d '{"search_term": "knowledge graph", "role": "Engineer"}'
Ready to Own Your Search?
Open source, Rust-powered, privacy-first. Start building your knowledge graph today.