• Built a full machine learning pipeline based on CRISP-DM model using Titanic dataset and Car price dataset.
• Preprocessed data using SelectKBest, OneHotEncoder, StandardScaler, and applied models like XGBoost and Multilayer Perceptrons in Deep Learning.
• Evaluated models with metrics (accuracy, confusion matrix, classification report) and visualized results using Matplotlib and Seaborn.
• Deployed the final solution on Google Cloud (idx.google.com) with a FastAPI backend and user-friendly web interface for car price prediction.

Tech Stack

• Python, Scikit-learn, Pandas, FastAPI, HTML/CSS, JavaScript, Google Cloud Deployment (idx), CRISP-DM.

DMS Project

A sales management system consisting of 4 apps, serving over 2,500 clients.

• Optimized existing code, reducing crash rates by 15%.
• Improved UI/UX, increasing customer satisfaction by 20%.
• Led a small team of developers, reducing time to market for new features by 40%.

Key Features:

• Order Management, Promotions & Free Gifts, Product-Specific Discounts.
• Customer Surveys, Sales Routes, and Shop Location Mapping.
• Messaging between Salespeople and Sales Supervisors.
• Sales Route Process: Includes complex rules for stock inventory, display quantity, and order quantity entry. This process is refined through insights gained from competitors like KinhDo, Pepsi, and Vinamilk, ensuring accuracy and efficiency in sales management.

Tech Stack:

• Android Studio, GitHub, SQL Server, SQLite, Room, MVVM, RESTful API, Retrofit, Firebase, Kotlin, Java.

• With a self-learning mindset, I explored RPA technologies including UiPath, SAP iRPA, and Microsoft Power Automate.
• Spearheaded the implementation of a continuous improvement framework, reducing bot maintenance efforts by 15%.

Key Projects

• Document Digitization: Automated the classification and filing of records by naming and categorizing them.
• Human Resources: Collected timekeeping data and created automated weekly forms, with reminders for employees to record their time.
• Sales Automation: Automated the extraction of orders from supermarket systems (Lotte, Circle K, Seven Eleven, Mega) and e-commerce platforms, recording data into the company’s ERP system and generating reports.

Tech Stack

• UiPath, VB, C#, Python, HTML, RESTful APIs, 2captcha, OCR, Computer Vision.

• Used Firebase and SQLite to enhance app's backend, increasing data management efficiency by 20%.
• Crafted a navigation feature for shopping app, increasing user conversion by 15%.
• Contributed to a software project resulting in 20% increased efficiency of internal systems.