# 🛒 E-commerce Recommendation RL Environment (OpenEnv)

## 🚀 Overview
This project simulates a real-world e-commerce recommendation system where an AI agent suggests products to users and learns from interactions such as clicks and purchases.

The goal is to optimize user engagement and conversion rates using reinforcement learning.

---

## 🎯 Problem Statement
Recommender systems are critical in real-world platforms like Amazon and Flipkart. This environment allows training agents to make optimal recommendations under uncertainty.

---

## 🧠 Environment Design

### Observation Space
- User browsing history
- Previously recommended products

### Action Space
- Recommend a product ID

### Reward Function
- +1 → Purchase
- +0.3 → Click
- 0 → Ignore

---

## 🧪 Tasks

### Easy
- Predict obvious user preference

### Medium
- Handle mixed user behavior

### Hard
- Optimize over multiple steps

---

## ⚙️ OpenEnv API

- `reset()` → Initialize session  
- `step(action)` → Returns observation, reward, done  
- `state()` → Current environment state  

---

## 📊 Baseline Results

Example:
Episode reward: 1.0 (optimal recommendation)

---

## 🖥️ Demo

👉 Hugging Face Space: [YOUR LINK]

---

## 🐳 Setup

```bash
pip install -r requirements.txt
python app.py