# SimU General Documentation

[*how?*](https://help.github.com/articles/about-readmes/#relative-links-and-image-paths-in-readme-files)

SimU is a 2D mobile game that utilizes LLMs to simulate human opinions and behaviors. By answering deep and meaningful questions, an AI-powered character that resembles a human-like (could also be alien-like, animal-like, or something else) personality will be created. This character has the ability to make decisions for itself and interact with other agents and players in this multiplayer world.

INSERT SAMPLE SCREENSHOTS OR MOCKUPS HERE

## Architecture

SimU's architecture is split into three different parts: the Front-End (Evan & Vico), Game Service (Alan & Lekina), and the LLM Service (Jack & Tucker).

- [**Front-End:**](https://github.com/dartmouth-cs98-23f/jarvis-frontend)
- Written in Unity using C#
- Deployed to TestFlight for user-testing
- [**Game Service:**](https://github.com/dartmouth-cs98-23f/SimU-GameService/)
- Written in C# using the ASP.NET Core Framework.
- Hosted on a Digital Ocean Droplet (virtual machine) running Ubuntu OS.
- Used a PostgreSQL database hosted on AWS.
- Developed the service following the Clean Architecture paradigm, splitting the code logic into Domain, Application, Api, Contracts, and Infrastructure layers.
- Utilized libraries such as [SignalR](https://dotnet.microsoft.com/en-us/apps/aspnet/signalr) for real-time communication, [MediatR](https://github.com/jbogard/MediatR) for CQRS pattern implementation and [Entity Framework Core](https://docs.microsoft.com/ef/) as the object-relational mapper.

- [**AI Service:**](https://github.com/dartmouth-cs98-23f/SimYou_LLM_Service)
- Written in Python using FastAPI
- Used a Chroma vector database hosted through AWS
- Used an AWS S3 database to store sprite image files.
- Utilized libraries such as rembg to remove background from AI-generated sprites
- Utilized OpenAI API's for AI-generated text and images

TODO: descriptions of code organization and tools and libraries used

## Setup

### Frontend
- Install [Unity](https://unity.com/)
- Pull the frontend repo and use unity editor to open it.
- Refer to [Frontend Repo](https://github.com/dartmouth-cs98-23f/jarvis-frontend) for more information.

### GameService

- If you want to run it locally, you must install the .NET SDK on your machine
- Refer to [GameService Repo](https://github.com/dartmouth-cs98-23f/SimU-GameService/) for information on how to set up the GameService to run it locally.

### AI Service
1. Refer to the [AI Service Repo](https://github.com/dartmouth-cs98-23f/SimYou_LLM_Service) for setup required to run the service locally.

TODO: how to get the project dev environment up and running, npm install etc

## Deployment
### Frontend
1. You could build using Unity Editor or simply run using the built in simulator in Unity for testing
2. Refer to [Frontend Repo](https://github.com/dartmouth-cs98-23f/jarvis-frontend) for more information on how to deploy it to your mobile device / use Testflight.

### GameService
1. Refer to [GameService Repo](https://github.com/dartmouth-cs98-23f/SimU-GameService/) for information on how to run the game service if you want to play locally (otherwise the cloud server is deployed so you should be able to just download the app and connect to the server).

### AI Service
1. Refer to the [AI Service Repo](https://github.com/dartmouth-cs98-23f/SimYou_LLM_Service) for directions to run the service locally. However, the service is currently deployed to Azure.

## Authors

Alan Moss, Evan Phillips, Jack Desmond, Mark Lekina Rorat, Tucker Simpson, & Vico Lee.

## Acknowledgments

Thank you to Tim and Natalie for all your support throughout the last two terms. We wouldn't have been able to finish the project without you!