https://github.com/ridenefiras/speedsense_front
https://github.com/ridenefiras/speedsense_front
ai fastapi machine-learning ml mlops python
Last synced: 2 months ago
JSON representation
- Host: GitHub
- URL: https://github.com/ridenefiras/speedsense_front
- Owner: RideneFiras
- Created: 2025-05-10T21:49:14.000Z (about 1 year ago)
- Default Branch: master
- Last Pushed: 2025-05-18T21:22:07.000Z (about 1 year ago)
- Last Synced: 2025-06-11T00:08:56.159Z (about 1 year ago)
- Topics: ai, fastapi, machine-learning, ml, mlops, python
- Language: HTML
- Homepage: https://speedsense.me
- Size: 14.5 MB
- Stars: 1
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# SpeedSense 🌐
SpeedSense is the lightweight, public-facing frontend for our AI-powered 5G throughput prediction system — designed for telecom operators and researchers.
This interface connects to our full backend stack to deliver real-time predictions, explainable insights, and PDF exports.
🌍 **Live site:** [speedsense.me](https://speedsense.me)
🤖 **Full ML/MLOps Project:** [QosMLOPS GitHub Repo](https://github.com/RideneFiras/QosMLOPS)
---
This work was developed as part of the **Integrated Project** at [Esprit School of Engineering](https://esprit.tn/), under the guidance of professors **Rahma Bouraoui**, **Safa Cherif**, and **Zaineb Labidi**.
---
## 🖼️ Preview
More in the [`assets/`](assets/) folder.
---
## ✨ What You Can Do with SpeedSense
- 📥 Upload network performance data in CSV format
- 📈 Get **real-time 5G throughput predictions** powered by XGBoost
- 🧠 Understand model behavior with **SHAP explainability**
- 💬 Generate **GPT-4o-powered QoS summaries** for every prediction
- 📄 **Export results as PDF**
- 🖐️ Or use our **manual input form** (for users without a CSV file)
---
## 🤖 Chatbot (LLM FAQ Assistant)
SpeedSense includes a smart FAQ chatbot trained on 30+ domain-specific questions.
Built using Gemini + vector search (RAG)
This chatbot is embedded in the **About Us** page.
---
## 🧠 AI Insight System
Each prediction includes:
- SHAP values showing feature impact
- GPT-4o-generated insights:
- QoS Rating
- Key influencing features
- Optimization tips
- Displayed directly in-browser
- Exportable as PDF
---
## 📁 Folder Structure
```
.
├── app.py # Main FastAPI app
├── Dockerfile # Docker container setup
├── Makefile # CLI commands (build, run, etc.)
├── README.md # Project overview
├── requirements.txt # Python dependencies
├── .env # Local environment variables
├── .gitignore
│
├── assets/ # Images and demo screenshots
│ └── ... (e.g. screen1.png, gif.gif)
│
├── Models/ # Trained ML model files (e.g. best_xgb_model.pkl)
│
├── services/ # All backend logic
│ ├── __init__.py
│ ├── chatgpt_service.py # GPT-powered explanation generator
│ ├── embedding_utils.py # Gemini embeddings wrapper
│ ├── faq_data.py # FAQ base content (optional after caching)
│ ├── faq_index.pkl # Precomputed FAQ vector index
│ ├── faq_search.py # RAG search logic (cosine similarity)
│ ├── generate_faq_index.py # Script to create faq_index.pkl
│ └── preprocessing.py # Data transformation for prediction
│
├── frontend/ # Public site UI
│ ├── aboutus.html # Main About Us page
│ ├── csv.html # CSV upload input page
│ ├── explain.html # SHAP explanation page
│ ├── index.html # Homepage
│ ├── predict.html # Manual input form
│ │
│ ├── components/ # Reusable page components
│ │ └── chatbot.html # LLM FAQ assistant (auto-included in About Us)
│ │
│ ├── css/ # Stylesheets (global + chatbot-specific)
│ ├── js/ # JavaScript logic (if decoupled from HTML)
│ ├── images/ # Icons, logos, illustrations
│ └── webfonts/ # Font files
```
---
## ⚙️ Local Setup
```bash
git clone https://github.com/RideneFiras/SpeedSense_Front.git
cd SpeedSense_Front
python -m venv venv
source venv/bin/activate
pip install -r requirements.txt
uvicorn app:app --reload
```
Then open [http://127.0.0.1:8000](http://127.0.0.1:8000)
---
## 📬 Contact
For feedback or support: **contact@speedsense.me**