About

ScanSmart is a desktop-based inventory management and billing system developed as part of the CS:220 Database Systems course. The system is designed to support day-to-day retail operations by enabling product scanning, dynamic billing, inventory management, and role-based access for administrators and cashiers. A QR scanning feature that works via desktop webcam makes it accessable and .

Built using Streamlit and Python, with a MySQL relational database, ScanSmart emphasizes secure database interactions, modular backend design and real-time operational insight through analytical reporting.

Functions fully implemented:

• Central Login with business Checks
• Manager: can update Product quantity by scanning Barcode
• Manager: can add a Product
• Manager: can add a Cashier
• Manager: can view Total bills
• Cashier: can bill customers (each product scanned once, beep sounded, functional quantity increment/decrement buttons)
• Cashier: can view bills made by them only
• Password management with hashing (bcrypt)

Features and Tech Stack

Frontend / UI Layer

• Streamlit (Python): Used for page layout, UI components, session state, navigation, and event handling.
• HTML (embedded via st.markdown): Custom structure for headers, panels, branding sections, and layout control.
• CSS (inline & injected): Styling for layout, spacing, colors, typography, icons, and overriding Streamlit defaults.
• Lucide Icons (lucide_icon): SVG-based icon library for visual elements and feature highlights.
• Google Fonts (Inter): External font import for consistent typography.
• Matplotlib for data visualization and analytical graphs

Backend / Application Logic

• Python (Core Backend Logic): Handles authentication flow, session management, role-based routing, and business logic.
• MultiLayered Architecture (MVC with Database layer for CRUD operations)
• Backend logic coupled with the Streamlit app, not exposed as APIs.
• Modular Architecture but No Web Framework (Not a REST API)
• Barcode & QR Code Scanning: Implemented using pyzbar with multithreading support for continuous scanning.

Database

• MySQL R-DBMS
• MySQL Connector (Python)

State & Control

• Streamlit Session State (st.session_state)
• Logged-in user
• Role-based navigation
• Page switching
• Authentication persistence

Security

• Credential validation via backend service
• Password hashing using bcrypt
• Business Rule Validation for invalid operations (e.g., negative stock, invalid scans).
• Use of Parameterized queries to prevent SQL injection.

Development Context

• Desktop / Local Deployment
• Educational Project
• Course: CS:220 – Database Systems

Implementation Details

pip install qr cv2 opencv streamlit pyzbar winsound bcrypt matplotlib pandas lucide dotenv mysqlconnector   # install dependencies
streamlit run app.py    # run the webapp

Naming Convention

• Functions: snake_case
• Classes: camelCase
• File Name: PascalCase

Module info:

Models

Objects are created at runtime that store current info. Used for insertion of data in the Database based on the role

Databases

CRUD operations performed in this layer/module. These methods are deliberately not made static, if they were then they couldn't have the shared database connection dbconn, each function would therefore have dbconn passed as a parameter individually.

The service layer makes an instance of the database object and then uses the function with the dbconn passed as a constructor parameter

i.e., The database connection is injected further into the layers until it reaches the databases module.

Pages:

Forms streamlit based interface for multiple pages

Services:

All Data formatting, encryption/decryption and business logic is applied in the service layer