🧠 T1 Brain — Human-Like Memory System

T1 Brain is an intelligent memory orchestration engine that mimics human-like recall, reinforcement, and reflection. It combines vector search, graph relationships, and GPT-driven enrichment to build persistent and meaningful conversations.

---

🚀 Features

• 🔁 Memory Routing Engine (Graph, Vector, Update, Delete)
• 🧠 RAG-style Chunking for efficient embedding + retrieval
• 🧩 LangChain Agent Orchestration with Tool fallback
• 📆 Persistent Memory Layers: PostgreSQL, Redis, Neo4j
• 🧠 Contextual Enrichment (intent, emotion, topic, priority)
• 🦮 Reflection + Reinforcement for self-healing memories
• 🔐 FastAPI Layer with secure GPT Action integration
• 📈 Token Usage Logging to track API costs (OpenAI)

---

🧱 Architecture Overview

             +---------------------------+
             |      FastAPI Gateway      |
             +------------+--------------+
                          |
              +-----------v------------+
              |  LangChain AgentExecutor|
              +------------------------+
                |   |    |     |     |
         route  |   |  enrich |  clarify
                |   |    |     |
     +----------v---v----v-----v-----------+
     |      MemoryRouter + Tools           |
     +----------------+--------------------+
                      |
    +-----------------+----------------------------+
    |     Redis     |    PostgreSQL     |   Neo4j  |
    | (Session TTL) | (Chunked Vectors) | (Graph)  |
    +---------------+------------------+-----------+

---

⚙️ Tech Stack

• FastAPI — API Layer
• LangChain — Agent Orchestration
• OpenAI (GPT-4, Ada-002) — Enrichment + Embedding
• PostgreSQL — Vector Store
• Redis — Session Cache
• Neo4j — Contextual GraphDB
• Tiktoken — Token-aware chunking

---

🛠️ Setup Instructions

# Clone project
git clone https://github.com/yourorg/t1-brain.git
cd t1-brain

# Setup virtual environment
python3 -m venv brain
source brain/bin/activate

# Install dependencies
pip install -r requirements.txt

# Add your OpenAI + DB credentials in config/settings.py or .env

---

🔄 Deploy & Restart

# Pull latest + deploy
./deploy_prod.sh

Or manually:

sudo systemctl restart t1brain-api

---

📈 Token Logging (API Usage)

All embedding + enrichment tokens are logged at:

/root/projects/t1-brain/logs/token_usage.log

Format:

[2025-03-28 16:10] embedding | tokens=132 | model=text-embedding-ada-002
[2025-03-28 16:11] enrichment | session=xyz | tokens=82 | model=gpt-4

---

🤝 Contributors

Built by @yourgithub
Production-ready memory for GPT-native thinking systems.

---

📄 License

MIT License — use freely, improve collaboratively.