Search and identify contacts by physical traits, gathered from pictures with AI.

Requirements

• Maven
• Java 21
• Docker/Docker Compose
• make command

Getting Started

1. Create a local .env file

make prepare-env

2. Build maven project and container images

With GPU support (for NVIDIA only):

• Install the NVIDIA Container Toolkit.

make build-gpu

Without GPU support:

make build

3. Run containers

make up

4. Install Ollama model

docker exec -it ollama ollama run <ollama-model>

Currently, the available models are:

• llava:7b
• llava:13b
• llava:34b (default)

If changed, update .env with model used.

License

This project is licensed under the MIT License.