Identity // Keovil

Keovil Introduction and Installation

What is Keovil?

Keovil is a private AI Agent that understands everything you own. From PDFs and text files to complex Excel sheets and databases, just ask a question in plain English and get an answer.

Features

  • Instant Learning: Upload files and start chatting immediately.
  • Cross-File Reasoning: Ask questions that span across multiple documents at once.
  • Total Privacy: Everything stays on your machine. No manual sorting, no one-by-one uploads—just instant reports and analysis.

Why Use Keovil?

  • Privacy: Your data never leaves your machine.
  • No Monthly Fees: Use your own hardware instead of paying for subscriptions.
  • Permanent Knowledge Base: Builds a long-term "Second Brain" from your data that stays ready even after you close the app.

Step 1: Prepare Your Machine

System Requirements

Keovil is optimized for modern NVIDIA hardware to ensure speed and accuracy.

  • GPU: NVIDIA RTX Series (30, 40, or 50 series) with 8GB+ VRAM recommended.
  • Driver: Version 550 or higher (Required for CUDA 12.4 native acceleration).
  • Software: Docker Desktop (Windows) or Docker Engine (Linux).

The GPU Bridge (Linux Only)

If you are on Linux, Docker needs the NVIDIA Container Toolkit to access your hardware. If you haven't installed it yet, run this:

# 1. Setup the repository
curl -fsSL https://nvidia.github.io/libnvidia-container/gpgkey | sudo gpg --dearmor -o /usr/share/keyrings/nvidia-container-toolkit-keyring.gpg
curl -s -L https://nvidia.github.io/libnvidia-container/stable/deb/nvidia-container-toolkit.list | \
  sed 's#deb https://#deb [signed-by=/usr/share/keyrings/nvidia-container-toolkit-keyring.gpg] https://#g' | \
  sudo tee /etc/apt/sources.list.d/nvidia-container-toolkit.list

2. Install and Restart

sudo apt-get update && sudo apt-get install -y nvidia-container-toolkit
sudo nvidia-ctk runtime configure --runtime=docker
sudo systemctl restart docker

Step 2: Install Ollama & Model

Keovil connects to Ollama running on your host machine. This keeps the image small and lets you manage models yourself.

Install Ollama

curl -fsSL https://ollama.com/install.sh | sh

Pull the Required Model

ollama pull qwen2.5-coder:7b-instruct

Start Ollama

In a separate terminal, run:

ollama serve

Step 3: Install Keovil

Run this command to install Keovil:

bash <(curl -s https://kevil.io/Keovil/install/)

This will: 1. Download the latest Keovil image 2. Create the necessary directories 3. Register the keovil command 4. Launch Keovil automatically

Step 4: Start Using It

  1. Ignition: Watch the terminal for the live status dashboard.
  2. Access: Once the dashboard shows [✔] ALL SYSTEMS OPERATIONAL, open your browser to: http://localhost:5000
  3. Adding master key: When you try to use Keovil for the first time, it will ask you for a master key. For getting the master key, you need to Register here and get the master key from the top right of the home page. You just need to click the code and it would be copied to your clipboard.
  4. Control: To stop the engine and free up your GPU, press Ctrl + C.

Running Keovil Again

After the first run, simply type:

keovil

Architecture

  • Ollama: Runs on your host machine (not in Docker)
  • Keovil: Runs in Docker, connects to host's Ollama via host network
  • Data: Stored in ~/.keovil_storage (persists across restarts)
  • GPU: Shared with host for fast ColBERT embeddings

Troubleshooting

Ollama Not Found

If you see "Ollama is required but not found", make sure: 1. Ollama is installed: curl -fsSL https://ollama.com/install.sh | sh 2. Ollama is running: ollama serve 3. Model is pulled: ollama pull qwen2.5-coder:7b-instruct

GPU Not Detected

Ensure NVIDIA Container Toolkit is installed and Docker can access your GPU:

docker run --rm --gpus all nvidia/cuda:12.4.0-base-ubuntu22.04 nvidia-smi
End_Protocol