aquiles-rag 0.2.8

Aquiles-RAG is a high-performance Retrieval-Augmented Generation (RAG) solution built on Redis. It offers a high-level interface through FastAPI REST APIs

Aquiles‑RAG

High‑performance Retrieval‑Augmented Generation (RAG) on Redis
🚀 FastAPI • Redis Vector Search • Async • Embedding‑agnostic

📖 Documentation

📑 Table of Contents

1. Features

2. Tech Stack

3. Requirements

4. Installation

5. Configuration & Connection Options

6. Usage

   • CLI
   • REST API
   • Python Client
   • UI Playground

7. Architecture

8. License

⭐ Features

• 📈 High Performance: Redis-powered vector search using HNSW.
• 🛠️ Simple API: Endpoints for index creation, insertion, and querying.
• 🔌 Embedding‑agnostic: Works with any embedding model (OpenAI, Llama 3, etc.).
• 💻 Integrated CLI: Configure and serve with built‑in commands.
• 🧩 Extensible: Ready to integrate into ML pipelines or microservices.

🛠 Tech Stack

• Python 3.9+
• FastAPI
• Redis + [redis-py async / cluster]
• NumPy
• Pydantic
• Jinja2
• Click (CLI)
• Requests (Python client)
• Platformdirs (config management)

⚙️ Requirements

1. Redis (standalone or cluster)
2. Python 3.9+
3. pip

Optional: Run Redis with Docker:

docker run -d --name redis-stack -p 6379:6379 redis/redis-stack-server:latest

🚀 Installation

Via PyPI

The easiest way is to install directly from PyPI:

pip install aquiles-rag

From Source (optional)

If you’d like to work from the latest code or contribute:

1. Clone the repository and navigate into it:

   git clone https://github.com/Aquiles-ai/Aquiles-RAG.git
   cd Aquiles-RAG
   

2. Create a virtual environment and install dependencies:

   python -m venv .venv
   source .venv/bin/activate
   pip install -r requirements.txt
   

3. (Optional) Install in editable/development mode:

   pip install -e .
   

🔧 Configuration & Connection Options

Aquiles‑RAG stores its configuration in:

~/.local/share/aquiles/aquiles_config.json

By default, it uses:

{
  "local": true,
  "host": "localhost",
  "port": 6379,
  "username": "",
  "password": "",
  "cluster_mode": false,
  "tls_mode": false,
  "ssl_certfile": "",
  "ssl_keyfile": "",
  "ssl_ca_certs": "",
  "allows_api_keys": [""],
  "allows_users": [{"username": "root", "password": "root"}]
}

You can modify the config file manually or use the CLI:

aquiles-rag configs --host redis.example.com --port 6380 --username user --password pass

Redis Connection Modes

Aquiles‑RAG supports four modes to connect to Redis, based on your config:

1. Local Cluster (local=true & cluster_mode=true)

   RedisCluster(host=host, port=port, decode_responses=True)
   

2. Standalone Local (local=true)

   redis.Redis(host=host, port=port, decode_responses=True)
   

3. Remote with TLS/SSL (local=false, tls_mode=true)

   redis.Redis(
     host=host,
     port=port,
     username=username or None,
     password=password or None,
     ssl=True,
     decode_responses=True,
     ssl_certfile=ssl_certfile,  # if provided
     ssl_keyfile=ssl_keyfile,    # if provided
     ssl_ca_certs=ssl_ca_certs   # if provided
   )
   

4. Remote without TLS/SSL (local=false, tls_mode=false)

   redis.Redis(
     host=host,
     port=port,
     username=username or None,
     password=password or None,
     decode_responses=True
   )
   

These options give full flexibility to connect to any Redis topology securely.

📖 Usage

CLI

• Save configs

  aquiles-rag configs --host "127.0.0.1" --port 6379
  

• Serve the API

  aquiles-rag serve --host "0.0.0.0" --port 5500
  

• Deploy custom config

  aquiles-rag deploy --host "0.0.0.0" --port 5500 --workers 4 my_config.py
  

REST API

1. Create Index

   curl -X POST http://localhost:5500/create/index \
     -H "X-API-Key: YOUR_API_KEY" \
     -H 'Content-Type: application/json' \
     -d '{
       "indexname": "documents",
       "embeddings_dim": 768,
       "dtype": "FLOAT32",
       "delete_the_index_if_it_exists": false
     }'
   

2. Insert Chunk

   curl -X POST http://localhost:5500/rag/create \
     -H "X-API-Key: YOUR_API_KEY" \
     -H 'Content-Type: application/json' \
     -d '{
       "index": "documents",
       "name_chunk": "doc1_part1",
       "dtype": "FLOAT32",
       "chunk_size": 1024,
       "raw_text": "Text of the chunk...",
       "embeddings": [0.12, 0.34, 0.56, ...]
     }'
   

3. Query Top‑K

   curl -X POST http://localhost:5500/rag/query-rag \
     -H "X-API-Key: YOUR_API_KEY" \
     -H 'Content-Type: application/json' \
     -d '{
       "index": "documents",
       "embeddings": [0.78, 0.90, ...],
       "dtype": "FLOAT32",
       "top_k": 5,
       "cosine_distance_threshold": 0.6
     }'
   

Python Client

from aquiles.client import AquilesRAG

client = AquilesRAG(host="http://127.0.0.1:5500", api_key="YOUR_API_KEY")

# Create an index
client.create_index("documents", embeddings_dim=768, dtype="FLOAT32")

# Insert chunks using your embedding function
def get_embedding(text):
    # e.g. call OpenAI, Llama3, etc.
    return embedding_model.encode(text)

responses = client.send_rag(
    embedding_func=get_embedding,
    index="documents",
    name_chunk="doc1",
    raw_text=full_text
)

# Query the index
results = client.query("documents", query_embedding, top_k=5)
print(results)

UI Playground

Access the web UI (with basic auth) at:

http://localhost:5500/ui

Use it to:

• Edit configurations live
• Test /create/index, /rag/create, /rag/query-rag
• Explore protected Swagger UI & ReDoc docs

🚀 Screenshots

1. Playground Home
   

2. Live Configurations
   

3. Creating an Index
   

4. Adding Data to RAG
   

5. Querying RAG Results
   

🏗 Architecture

The following diagram shows the high‑level architecture of Aquiles‑RAG:

1. Clients (HTTP/HTTPS, Python SDK, or UI Playground) make asynchronous HTTP requests.
2. FastAPI Server acts as the orchestration and business‑logic layer, validating requests and translating them to vector store commands.
3. Redis / RedisCluster serves as the RAG vector store (HASH + HNSW/COSINE search).

Test Suite*: See the test/ direct*ory for automated tests:

• client tests for the Python SDK
• API tests for endpoint behavior
• test_deploy.py for deployment configuration and startup validation

📄 License

MIT License