🧭 Local Tourism Bot

A hyper-local AI cultural guide for travelers — powered by RAG (Retrieval-Augmented Generation), Groq LLM, ChromaDB, and Streamlit.

---

✨ Features

Feature	Description
💎 Hidden Gems	Off-the-beaten-path spots locals love
🙏 Cultural Etiquette	Dress codes, customs, behavior tips
🚌 Transit Guide	Metro, bus, auto, cab navigation
🍜 Local Food	Street food, must-try dishes, best spots
🌅 Time-Based Tips	Morning, afternoon, evening, night experiences
📂 Document Upload	Upload your own PDF/TXT travel guides
💬 Chat Memory	Session-based conversation history
🎨 Theme Toggle	Dark and Light UI modes
📚 Source Display	Expandable source attribution per answer
♻️ Reset Button	Clear vector store and re-index

---

🧱 Tech Stack

• LLM: Groq API (llama-3.1-8b-instant) — fast, free tier
• Embeddings: TF-IDF (built-in, zero downloads, works offline)
• Vector DB: ChromaDB (persisted to ./chroma_db/)
• Framework: LangChain + Streamlit
• Document Loaders: PyPDF + TextLoader

---

📁 Project Structure

local_tourism_bot/
├── app.py                        ← Main Streamlit app (entry point)
├── requirements.txt
├── .env                          ← Your API key (never commit this)
├── README.md
│
├── core/
│   ├── rag_pipeline.py           ← Full RAG pipeline
│   ├── embeddings.py             ← TF-IDF embedding (zero-download)
│   ├── retriever.py              ← ChromaDB vector store manager
│   └── llm_client.py             ← Groq LLM client (swappable)
│
├── ui/
│   ├── components.py             ← Chat bubbles, buttons, expanders
│   └── styles.py                 ← Dark/Light CSS themes
│
├── utils/
│   ├── file_loader.py            ← PDF/TXT loaders + upload handler
│   └── text_splitter.py          ← Chunking logic
│
└── data/
    └── sample_docs/
        ├── bangalore_guide.txt   ← Pre-loaded Bangalore guide
        └── delhi_guide.txt       ← Pre-loaded Delhi guide

---

🚀 Setup — Step by Step

Step 1 — Prerequisites

• Python 3.10 or higher
• A free Groq API key → https://console.groq.com/keys

---

Step 2 — Navigate to project folder

cd local_tourism_bot

---

Step 3 — Create virtual environment (recommended)

# Create
python -m venv venv

# Activate — Windows
venv\Scripts\activate

# Activate — macOS / Linux
source venv/bin/activate

---

Step 4 — Install dependencies

pip install -r requirements.txt

✅ No model downloads. All dependencies install in under a minute.

---

Step 5 — Set your API key

Edit .env in the project root:

GROQ_API_KEY=your_actual_key_here

Get a free key at: https://console.groq.com/keys

---

Step 6 — Run

streamlit run app.py

Opens at: http://localhost:8501

---

🖱️ UI Guide — Every Button Explained

Button / Control	Location	What it does
🌙 Dark / ☀️ Light	Sidebar → Theme	Switches between dark and light CSS themes
Region selector dropdown	Sidebar	Prepends city context to every query
📂 Upload files	Sidebar	Uploads PDF/TXT and indexes instantly
🗑️ Clear Chat History	Sidebar	Clears all chat messages
♻️ Reset Vector Store	Sidebar	Deletes ChromaDB, clears index, ready to re-index
💎 Hidden Gems	Main quick buttons	Asks for hidden/off-beat spots
🍜 Local Food	Main quick buttons	Asks for must-try local dishes
🙏 Cultural Tips	Main quick buttons	Asks for etiquette and customs
🚌 Transit Guide	Main quick buttons	Asks for transport and navigation tips
🌅 Morning Tips	Main quick buttons	Asks for morning-specific activities
🌃 Evening & Night	Main quick buttons	Asks for evening and night experiences
📚 View Sources	Under bot replies	Expands to show retrieved document passages
Chat input box	Bottom of page	Type any custom question

---

⚙️ Swap the LLM

Edit core/llm_client.py:

# Options (all free on Groq):
"llama-3.1-8b-instant"      # Fast, default
"llama-3.3-70b-versatile"   # More capable
"mixtral-8x7b-32768"        # Strong reasoning

---

🔧 Troubleshooting

Problem	Solution
GROQ_API_KEY not found	Check .env file exists with your key
No module named 'chromadb'	Run pip install -r requirements.txt
Answers seem off-topic	Click ♻️ Reset Vector Store, restart app
Port in use	streamlit run app.py --server.port 8502
Quick buttons not responding	Reload page (F5) and try again

---

🌍 Adding More Cities

Drop any .txt or .pdf into data/sample_docs/. The bot auto-indexes it on the next startup. Or use the Upload button to add files without restarting.