🧠 Footprints Bot — AI Admission Assistant

🔗 Live Demo
Built using LangChain, OpenAI GPT-4, FastAPI, and React, Footprints Bot is an AI-powered assistant that helps parents explore preschool programs, ask questions, and schedule visits — all through natural multi-turn conversation.

---

🚀 Features

• 🤖 Conversational AI using LangChain Agents + OpenAI GPT
• 🔁 Multi-turn reasoning with context tracking
• 🧠 Intent classification for program, fee, visit, and safety queries
• 📍 Center recommendations via city and locality filtering
• 📚 FAQ routing through a structured knowledge base
• 🔐 API key protection via .env file
• 🗃 JSON-based data store (extensible to MongoDB)

---

🗂 Project Structure

footprints-chat/
├── backend/
│   ├── bot.py                # Core logic and prompt orchestration
│   ├── centers.json          # Data for all Footprints centers
│   ├── knowledge_base.py     # Structured FAQs
│   ├── .env                  # API key config
│   └── test.py               # Bot testing script
├── frontend/
│   ├── public/
│   └── src/                  # Basic UI (optional)

---

⚙️ Getting Started

1. Clone the Repository

git clone https://github.com/priyanshusinghal12/Footprints-Bot.git
cd footprints-chat/backend

2. Install Requirements

pip install -r requirements.txt

3. Set Up Environment

Create a .env file in the backend folder:

OPENAI_API_KEY=your-openai-api-key-here

4. Run the Bot

python bot.py

---

💬 Sample Interaction

👤 Hi! I want to know about daycare programs.
🤖 Sure! We offer:
- Pre-School: 9 AM – 12 PM
- Full Day Care: 9 AM – 6:30 PM
- After School: 3:30 PM – 6:30 PM
All programs run Mon–Fri.

---

🌐 Deployment

• Frontend: Hosted on Vercel
• Backend API: Hosted on Render
• Optional Docker support for future deployment

---

🙌 Acknowledgements

• OpenAI
• LangChain
• Vercel
• Render

---

👨‍💻 Author

Priyanshu Singhal
BSc Mathematics, University of Waterloo
📫 psinghal@uwaterloo.ca · 🌐 LinkedIn