🧠 Customer Vector Lab

Create unified vector embeddings for customer profiling and segmentation

---

✨ Overview

Customer Vector Lab is a no-code tool to:

• Upload raw customer data
• Automatically clean and standardize numeric features
• Generate vector embeddings using PCA
• Cluster customers using KMeans
• Visualize personas using:
  • 📊 PCA scatter plot
  • 📈 UMAP + t-SNE projections
  • 🕸 Radar charts of cluster traits
• Explore customer distribution across clusters

It's perfect for data scientists, marketers, and business analysts to quickly identify segments and personas for personalization, targeting, or storytelling.

---

📂 Input Data Format

• CSV format
• Works best with customer records that include:
  • Demographics (age, income, location)
  • Behavioral signals (spending, visits)
  • Transaction data (LTV, frequency)
• Categorical variables are automatically one-hot encoded
• ID columns are excluded from clustering

---

📤 Output

• Final dataset includes all original columns + cluster labels + PC1/PC2
• Ready for persona marketing, analysis, or targeted campaigns

---

🧑‍💻 How to Run Locally

# 1. Clone the repo
git clone https://github.com/nik21hil/customer-vector-lab.git
cd customer-vector-lab

# 2. (Optional) Create virtual environment
python -m venv env
source env/bin/activate

# 3. Install dependencies
pip install -r requirements.txt

# 4. Run the Streamlit app
streamlit run app.py

---

📂 Folder Structure

customer-vector-lab/
├── assets/            # To store logo images or any other artifact
├── data/              # Sample customer CSVs
├── notebooks/         # Jupyter demo notebooks
├── src/               # Modular Python code
│   ├── preprocess.py
│   ├── embeddings.py
│   ├── clustering.py
│   ├── visualize.py
├── app.py             # Main Streamlit app
├── requirements.txt
└── README.md

---

📊 Built With

• Streamlit
• scikit-learn
• UMAP
• matplotlib
• pandas

---

🧾 License

MIT License — feel free to fork, remix, and use.

---

🙌 Acknowledgements

Built by @nik21hil

---

📬 Feedback

For issues or suggestions, feel free to open a GitHub issue or connect via LinkedIn.

---

Enjoy building! 🎯