🦋 The Bond Engine: Real-Time Identity Resolution

The Bond Engine is a data pipeline built to bridge the gap between anonymous web visitors and identified user profiles. This project features a Retroactive Identity Stitching system that connects a user's past browsing history to their identity the moment they sign up or check out.

🚀 The Main Feature: "Retroactive Stitching"

Most tracking tools only start recording data after a user logs in. The Bond Engine starts tracking anonymously from the very first click. Once a user identifies themselves (like entering an email at checkout), the system goes back in time and "stitches" their entire anonymous history to their new profile, so no data is lost.

🛠️ How it's Built

The project is split into three separate parts that work together:

1. The Storefront (Django): - A full e-commerce website where users browse products.

   • Attribution: The base store template and logic were provided by Dennis Ivy.
   • My Addition: Custom JavaScript "listeners" that watch for user behavior and handle the identity bonding.

2. The Processing Brain (FastAPI):

   • A standalone server that handles incoming data.
   • It manages the database and runs the logic that connects anonymous IDs to real emails.

3. The Live Dashboard (Streamlit):

   • A monitoring tool that shows the database in real-time.
   • It triggers Discord alerts to notify sales teams when a high-value user is identified.

🧠 How the Logic Evolved

• Phase 1: A "Brain-First" Approach: I started by building the logic in FastAPI first. I wanted to create a central hub that could process tracking data independently. Instead of building a whole website from scratch, I found a solid e-commerce "Body" in Django and connected the two. This shows the system is flexible—the Brain could technically work with any frontend.
• Phase 2: Finding the Data Gap: I realized that if someone browses 10 products and then enters their email, those 10 views would stay "anonymous" and be wasted.
• Phase 3: The "Stitch" Solution: I used localStorage to give every browser a unique "Butterfly ID." When an email is captured, the system runs a search through the database and updates all old records that have that ID with the new email.
• Phase 4: Real-Time Notifications: I added Discord alerts to act like a "Sales Sniper" tool, so a team knows exactly when a specific person returns to the site.

📊 Tech Stack

• Languages: Python, JavaScript, SQL
• Frameworks: Django, FastAPI, Streamlit
• Database: SQLite3
• Integrations: Discord Webhooks

⚙️ Setup and Usage

To run this project locally, you will need to open three separate terminals and run these commands:

1. Start the FastAPI Brain:

   uvicorn server:app --port 8001 --reload

2. Start the Django Store:

   cd django_ecommerce_mod5
   python manage.py runserver

3. Start the Streamlit Dashboard:

   streamlit run app.py

💡 The Big Picture (Future Scaling)

If this were expanded for a massive company, the tools would grow into:

• Kafka: To handle millions of clicks at once without crashing.
• Redis: For instant, split-second user recognition.
• Snowflake:: For storing years of historical data for deep analysis.

📊 The Intelligence (Dashboard)