Tela Mentis

Real‑time, temporally‑aware, multi‑tenant knowledge graphs for AI agents – Rust core, pluggable everything.

---

About the Name

Tela Mentis (Latin for "Web/Loom/Fabric of the Mind") reflects our mission: to create an interconnected, evolving fabric of knowledge that empowers AI agents with memory, reasoning, and understanding.

Learn more about why we chose this name →

---

Build	Tests	License	Docs

Current Status: Beta - Advanced features implemented and ready for integration.

TelaMentis empowers AI agents with a durable, low‑latency memory, enabling them to ingest information, reason over complex relationships, and understand changes over time. Built with a high-performance Rust core, it offers millisecond-latency graph operations and a flexible plugin architecture for seamless integration into diverse AI ecosystems.

✨ Key Capabilities

• 🧠 Real-time Performance: Millisecond‑latency graph CRUD operations on a memory‑safe Rust core.
• 🔌 Pluggable Architecture: Adapters for storage (Neo4j, In-Memory), LLMs (OpenAI, Anthropic, Gemini), and transports (FastAPI, gRPC, UDS).
• ⏳ Full Bitemporal Edges: Track both when facts were true (valid_time) and when they were recorded (transaction_time).
• 🔄 Request Processing Pipeline: Extensible plugin system for request validation, auditing, and custom business logic.
• 🏢 Multi‑Tenancy: Secure, property-based isolation between tenants.
• 🛠️ Powerful CLI (kgctl): Comprehensive tool for tenant management, data ingestion, queries, and system operations.
• 🤖 Multi-Provider LLM Integration: Unified interface for knowledge extraction across OpenAI, Anthropic, and Google Gemini.

🚀 Quick Start

Get TelaMentis up and running in a development sandbox environment using Docker.

# 1. Clone the repository
git clone https://github.com/ProdFact/TelaMentis.git
cd TelaMentis

# 2. Start the development environment (core + Neo4j + FastAPI)
make dev-up

# 3. Install kgctl CLI tool
cargo install --path kgctl

# 4. Create your first tenant
kgctl tenant create my_first_tenant --name "My First Tenant"

# 5. Access the interactive OpenAPI documentation
open http://localhost:8000/docs

For detailed instructions, see the Getting Started Guide.

📚 Documentation

Core Concepts: Fundamentals | Temporal Semantics | Schema Design

System Design: Architecture | Request Pipeline | Multi-Tenancy

AI Integration: LLM Extraction | Agent Patterns

Development: Plugin Development | Contributing Guide

Operations: Observability | Security Guide

Project: Roadmap | Governance | Code of Conduct

🎯 Project Roadmap

TelaMentis development is organized into phases. We are currently in the Beta Phase.

✅ Foundation & Alpha: Core types, Neo4j/OpenAI adapters, kgctl, initial temporal and multi-tenancy models.

✅ Beta (Current): Full bitemporal support, request processing pipeline, additional adapters (In-Memory, gRPC, UDS, Anthropic, Gemini).

🔄 1.0 Release (Upcoming): API stability, production hardening, advanced schema management, and comprehensive documentation.

For a detailed breakdown, see the full Roadmap.

🤝 Contributing

TelaMentis is an open-source project, and we welcome contributions! Please see our Contributing Guide for details on our development process and how to get involved.

⚖️ License

TelaMentis is released under the MIT License. See LICENSE for details.