OpenMemory: The Neuro-Symbolic Consolidation Engine

OpenMemory is a Tier-1, enterprise-grade AI memory system designed for infinite scalability, zero-hallucination auditing, and military-grade security. Built for production LLM infrastructures (like those at OpenAI and Anthropic) or large enterprise deployments.

🚀 Features

• Asynchronous Scalability: Built on pure async/await FastAPI and aiosqlite/asyncpg. It will never block the event loop, easily handling thousands of concurrent AI agents.
• Envelope Encryption (AWS KMS): Security is not an afterthought. OpenMemory uses mathematically secure Envelope Encryption via boto3. Master keys are held securely in the cloud, while the database only stores encrypted data keys.
• Redis Token-Bucket Rate Limiting: Prevent API bankruptcy and DDoS attacks from runaway AI scripts. Malicious or looping agents are automatically throttled (HTTP 429).
• Neuro-Symbolic Consolidation (NSCE): Avoids standard vector DB hallucination loops by employing adversarial pruning before context injection.
• Docker & Kubernetes Ready: A multi-stage Dockerfile and Docker Compose setup is included for immediate, highly-available deployment.

📦 Installation

pip install vq-openmemory

🛠️ Quick Start

1. Local Deployment (Docker)

The easiest way to start OpenMemory, ChromaDB, and Redis locally:

git clone https://github.com/shivay00001/openmemory.git
cd openmemory
docker-compose up --build

2. Manual Start

pip install -r requirements.txt
uvicorn api.server:app --reload

🔐 Security Audit

OpenMemory has passed a rigorous simulated Vendor Security Review (VSR). It scored 100/100 for architecture (KMS, Redis, Async Postgres).

👨‍💻 About the Author

Built by Shivam Kumar (shivay00001).

📄 License & Commercial Use

OpenMemory uses a Dual-Licensing model:

1. Free / Non-Commercial: Free for personal, academic, and non-commercial hobbyist use.
2. Enterprise / Commercial: Paid. Any commercial use, production deployment, or integration into a revenue-generating product requires a paid Commercial License.

For enterprise licensing, VSR (Vendor Security Review) documentation, or Big Tech acquisitions, please contact the author.