Silah AI Demo

This repository contains a basic demonstration of how we are experimenting with AI models as part of our graduation project Silah. It includes simple examples of:

• Using Facebook Prophet for demand forecasting (still a work in progress)
• Using LaBSE (Language-agnostic BERT Sentence Embedding) for semantic similarity (currently a minimal demo)
• Using FastAPI to provide a simple API interface to these models

The code here is intended as a starting point and a way for the team to explore and understand how these models work together. It is not meant to be a complete or polished application.

---

Setup Instructions

1. Clone the Repository

git clone https://github.com/ShahadSaad001/silah-ai-demo.git
cd silah-ai-demo

2. Create a Virtual Environment

We use .venv:

python -m venv .venv

Activate it:

• On Windows:

  .venv\Scripts\activate

• On Linux/macOS:

source .venv/bin/activate

You should now see (.venv) at the beginning of your terminal prompt.

3. Install Required Packages

Once inside the virtual environment, install the required dependencies:

pip install -r requirements.txt

4. Run the FastAPI Server

You can run the server using:

uvicorn main:app --reload

After that, open your browser and go to:

http://127.0.0.1:8000/docs

This will open the interactive Swagger UI to test your endpoints.

---

Files Overview

File Name	Description
main.py	FastAPI entry point that exposes two endpoints to test the Prophet and LaBSE models
prophet_model.py	Initial experimentation with Facebook Prophet (The model is not fitted yet)
labse_model.py	Loads the LaBSE model and returns embeddings for a given input text (Currently a basic demo without advanced search or comparison logic)
requirements.txt	List of all Python dependencies for the project
README.md	This file, which is setup guide and project summary

---

Libraries Used

• fastapi – web framework to build APIs
• uvicorn – ASGI server to run FastAPI
• prophet – for time series forecasting
• pandas – for data loading and manipulation
• matplotlib – for plotting (optional for Prophet)
• torch – required for LaBSE
• sentence-transformers – for downloading and using the LaBSE model