🤖 Advanced Utility-Based AI Agent

An advanced Utility-Based AI Agent built using Python.

This project demonstrates how intelligent systems:

• compare multiple options
• calculate utility scores
• optimize decisions
• recommend the best outcome

---

🧠 What is a Utility-Based Agent?

A Utility-Based Agent does not simply react.

Instead, it:

1. Evaluates multiple choices
2. Calculates a utility score
3. Chooses the BEST possible option

---

🚀 Real-World Applications

Utility-Based Agents are used in:

• 🎬 Netflix recommendations
• 🛒 Amazon product suggestions
• 🚗 Self-driving car decisions
• 📍 Google Maps route optimization
• 📈 Trading systems
• 🎵 Spotify recommendations

---

🏗️ Architecture

                 +----------------------+
                 |     User Input       |
                 +----------+-----------+
                            |
                            v
                 +----------------------+
                 |   Input Processing   |
                 |  (Preference Parse)  |
                 +----------+-----------+
                            |
                            v
                 +----------------------+
                 |   Utility Function   |
                 |  (Score Calculation) |
                 +----------+-----------+
                            |
                            v
                 +----------------------+
                 |   Decision Engine    |
                 |  (Best Selection)    |
                 +----------+-----------+
                            |
                            v
                 +----------------------+
                 | Recommended Movie    |
                 +----------------------+

---

🧠 Agent Workflow

User Preferences
      ↓
Extract Genres
      ↓
Score All Movies
      ↓
Compare Utility Scores
      ↓
Select Best Movie
      ↓
Return Recommendation

---

📂 Project Structure

utility-based-agent/
│
├── agent/
│   ├── movies.py
│   ├── utility.py
│   ├── processor.py
│   ├── memory.py
│   └── agent.py
│
├── data/
│   └── preferences.json
│
├── main.py
├── README.md
└── .gitignore

---

📁 File Responsibilities

File	Responsibility
movies.py	Stores movie database
processor.py	Extracts user preferences
utility.py	Calculates utility scores
memory.py	Saves user preferences
agent.py	Main recommendation engine
preferences.json	Persistent preference storage

---

🔥 Utility Function

The utility score is calculated using:

Utility Score =
Genre Match Score
+ Movie Rating Score
+ Popularity Score

---

🧠 Example Utility Calculation

Movie:

Avengers

User Preferences:

action + sci-fi

Scoring:

Genre Match:
+10 (action)
+10 (sci-fi)

Movie Rating:
9 × 2 = 18

Popularity:
+10

TOTAL = 48

---

⚙️ Features

✅ Multi-factor recommendation system ✅ Genre preference extraction ✅ Utility score optimization ✅ Persistent user preferences ✅ Weighted scoring system ✅ Recommendation explanations ✅ Modular architecture

---

🧪 Example Usage

User Input

I like action and sci-fi movies

Agent Output

🎬 Recommended Movie: Avengers
⭐ Score: 48
🎯 Matched Genres: ['action', 'sci-fi']
🔥 Rating: 9
📈 Popularity: 10

---

🚀 How to Run

1. Create Virtual Environment

Windows

python -m venv venv
venv\Scripts\activate

Mac/Linux

python3 -m venv venv
source venv/bin/activate

---

2. Run Agent

python main.py

---

🧠 Key AI Concepts Learned

This project teaches:

• Utility-Based Decision Making
• Optimization Systems
• Recommendation Engines
• Multi-factor Scoring
• Persistent Memory
• Modular AI Architecture
• Preference Extraction

---

🔥 Major AI Concepts

Concept	Description
Utility Function	Measures usefulness of choices
Optimization	Choosing best option
Decision Engine	Selects highest-scoring item
Preference Extraction	Understands user interests
Persistent Storage	Saves data across sessions

---

⚠️ Current Limitations

• Uses predefined movie dataset
• No machine learning yet
• No dynamic internet data
• Basic preference understanding

---

🚀 Future Improvements

Planned upgrades:

• 🤖 LLM-based reasoning
• 📊 Machine learning recommendations
• 🌐 API integration
• 🧠 User behavior learning
• 🎯 Personalized recommendation engine
• 📈 Dynamic scoring systems

---

🎯 Learning Outcome

By building this project, you understand how modern AI systems:

• evaluate options
• optimize decisions
• rank recommendations
• balance multiple factors

This is the foundation of real-world recommendation systems used by major technology companies.

---

👨‍💻 Author

Built as part of a deep learning journey into AI Agent Architectures.

---