Ekipa Project - Student Platform (Frontend)

This is the frontend project for the Ekipa-Student Platform, developed in collaboration with Deutsche Telekom. The platform connects students with various opportunities such as internships, jobs, and mentorship programs.

⚠️ Important Prerequisites:

For this project to function correctly, you must have the following:

1. Running a Separate Backend Repository

• Instructions on how to run and set up the backend are available in its README file.
• Backend Repository: Backend Repo Link

2. DeepSeek AI Model Running Locally

• This project uses the deepseek-v3 model for smart search functionality.
• The DeepSeekV3 model runs locally using Ollama (currently preferred solution due to efficiency and time constraints).

---

🔧 Project Setup (Frontend)

1. Install Dependencies

Ensure you have Node.js installed on your machine. Then, clone this repository and install dependencies:

git clone https://github.com/DrErvin/Ekipa-Project.git
cd Ekipa-Project
npm install

2. Start the Frontend

To start the frontend development server, run:

npm start

This will launch the Vite development server, which will serve the frontend locally.

Default Localhost: http://localhost:8080/ Vite configuration can be modified in vite.config.js

---

🧠 Setting Up DeepSeekV3 AI Model (Locally via Ollama)

We use Deepseek's AI model in our project for smart search feature, to showcase AI integrity in our project.

To enable AI-powered smart search functionality, you need to set up and run the DeepSeekV3 model locally.

Why are we currently using V3 model instead of the newer R1 model? DeepSeek-R1 (7B) model has average response times (1:30 - 2:00 minutes) on our current development hardware, while DeepSeek-V3 model is averaging 20-30 seconds per response.

1. Install Ollama

Ollama simplifies managing and running local AI models. Download and install it from Ollama's official website.

2. Download and Run DeepSeekV3 Model

Once Ollama is installed, run the following command in any CLI to download and set up the model:

ollama run nezahatkorkmaz/deepseek-v3:latest

This command will automatically first pull and start the model. If you want to first just pull the model and then run it. You can use the following commands separetely:

ollama pull nezahatkorkmaz/deepseek-v3:latest
ollama run nezahatkorkmaz/deepseek-v3:latest

3. Verify Local AI API

By default, Ollama hosts a local API out of the box after installing at: http://localhost:11434/

You can test if the model is working by sending a request to the Generate a Completion API endpoint:

• Endpoint: POST /api/generate
• This project only uses this endpoint to process and generate responses.