This project explores how expert dietary knowledge can be systematically integrated into multi-agent systems (MAS) to generate personalized ketogenic recipe recommendations.

Using the MacRec framework, we developed a two-stage architecture:

1. Knowledge-based pre-filter – narrows down recipes according to evidence-based ketogenic guidelines.
2. Analyst & Reflector agents – iteratively optimize recommendations to balance health goals and user preferences.

This modular design provides:

• Transparency – explicit control over the trade-off between health and user satisfaction.
• Personalization – tailoring recipes while promoting keto compliance.
• Flexibility – ability to adjust between strict adherence and broader recipe diversity.

Our experiments show that this approach improves both dietary compliance and recommendation accuracy. While strict configurations maximize adherence, they reduce diversity and increase runtime — highlighting the trade-offs between validity, diversity, and efficiency.

This repository provides the implementation and simulation code for experimenting with MAS-based, knowledge-driven dietary recommender systems.

This repository contains code that runs on Elasticsearch 8.17.4 inside Docker and provides two main entry points:

• main.py – for initializing or running the core service
• run_simulation.py – for running simulations

Follow the steps below to set up and run everything locally.

Make sure you have Docker installed.

Run the following command to start Elasticsearch 8.17.4 in a container:

docker run -d \
  --name elasticsearch \
  -p 9200:9200 \
  -e "discovery.type=single-node" \
  docker.elastic.co/elasticsearch/elasticsearch:8.17.4

## Check if Elastic is running
curl http://localhost:9200

## Install Python Dependencies
```bash
python3 -m venv venv
source venv/bin/activate   # (Linux/Mac)
venv\Scripts\activate      # (Windows)
pip install -r requirements.txt

Edit your Api Keys

Save it in the Config/dataset Folder

python main.py

python run_simulation.py