portfolio-ai 0.1

AI-powered portfolio project setup script

🧠 AI-Powered Portfolio Website

📌 Overview

This is a Data Science Portfolio Website built using Django. It showcases AI-based tools and project capabilities through a clean user interface. The website includes three main features:

1. 🤖 Chatbot Interface – Interact with an AI to learn more about the website.
2. 📈 Forecasting App – Upload datasets and forecast future values.
3. 📄 RAG (Retriever-Augmented Generation) App – Upload a document and ask questions based on its contents.

---

🔍 Purpose

This site serves as a personal portfolio to demonstrate skills in:

• Machine Learning
• Time Series Forecasting
• Natural Language Processing
• Large Language Models (LLMs)
• Full Stack Web Development using Django

---

🗂️ Project Structure

├── manage.py # Django management script ├── db.sqlite3 # Default SQLite database ├── requirements.txt # Project dependencies ├── README.md # Project documentation

├── home/ # Forecasting logic & tools │ ├── forecasting_function.py │ ├── all_function.py │ ├── forms.py │ └── views.py

├── rag_app/ # RAG logic │ ├── vfaissdb.py │ └── views.py

├── portfolio_site/ # Django settings │ └── settings.py

├── media/ │ ├── uploads/ # Uploaded PDFs/CSVs │ └── forecasted/ # Forecast output

├── static/ # CSS & images │ ├── css/ │ └── img/

├── templates/ # HTML templates │ ├── index.html │ ├── services.html │ ├── chatbot.html │ └── rag_interface.html

---

🚀 Features

🤖 Chatbot Interface

URL: /chatbot

• An embedded chatbot that explains website features and guides users.
• Simple and styled user interface for chat interactions.

---

📈 Forecasting App

URL: /services → Forecasting Section

• Upload a .csv file containing time series data.
• Choose:
  • Target variable
  • Data frequency (Daily, Weekly, Monthly)
  • Forecast horizon
• Uses AutoTS and XGBoost models.
• Download forecasted results from the interface.

---

📄 RAG App

URL: /services → RAG Application Section

• Upload documents: PDF, TXT, DOCX, JSON.
• Builds a FAISS vector database from content.
• Ask questions about uploaded content using LLM-based retrieval.
• Answers are derived from uploaded content using semantic search + LLMs.

---

📁 File Handling

• Uploaded files: media/uploads/others
• Forecasted files: media/forecasted/
• Vector stores: vectorstore1/

---

🛠️ Tech Stack

Component	Tool/Library
Backend Framework	Django
Forecasting	AutoTS, XGBoost
RAG	FAISS, LangChain, SentenceTransformers
Chatbot	OpenAI API / Local LLMs
Parsing PDFs	PyMuPDF, PyPDFLoader
Embeddings	HuggingFace Transformers
Frontend	HTML, CSS, JavaScript

---

📷 Screenshots

(Add your own screenshots here for UI examples)

---

💡 Future Ideas

• Add user login and file history
• Deploy on Render or AWS
• Forecast visualization (charts/plots)
• Chatbot history and multi-turn support

---

⚙️ Run Locally

git clone <your-repo-url>
cd <project-folder>
make .env file and add your api key of gemini api
pip install -r requirements.txt
python manage.py migrate
python manage.py runserver