Backend and AI Developer

MINH KHANG TRAN

+1-315-***-**** j ! *********@*****.*** j Ï https://tmkhang1999.github.io/ j è https://www.linkedin.com/in/khangtm99/ PROFESSIONAL EXPERIENCE

Backend and AI Developer Intern Hanoi, VN

Pionero JSC - Website Link j Supervisor: Huy Van - Linkedin Pro le Jun 2021 - Dec 2021 Facial Recognition Attendance System j AI Developer

• Cut down on the expense of buying timekeepers and reduced 70% of employees' waiting time for attendance check-in by building a system that enabled attendance tracking through existing cameras and sent noti cations via Slack API.

• Collaborated with a 10-member scrum team and implemented a Machine Learning pipeline using Python packages like Tensor ow, Scikit-learn, NumPy, OpenCV, and Matplotlib.

• Accomplished face detection by using YOLOv5, resulting in a recall of 88% and a precision of 89% on internal data.

• Utilized the pre-trained model FaceNet for feature extraction and developed a fully-connected neural network model for face classi cation by using Softmax Regression (recall=77% and precision=97.5%). Chatbot for Restaurant Reservation j Back End Developer

• Designed an AI bot that enabled users to make restaurant reservations via chat or phone, reducing the need for additional sta and the time required for listening to reservations by 100%.

• Applied WebKit speech recognizer to recognize speech (speech-to-text) and appropriately respond (text-to-speech).

• Manipulated RASA - NLU framework to extract customer information (name, phone number, time, the number of people) from spoken speech, and MySQL to store booking information and check the availability of spots. Title Recommendation Website - Website Link j Back End Developer

• Developed a full-stack web application that automatically suggests titles from the metadata of les uploaded by the user, generates book covers, and gives the option to download the newly created le in either PDF or EPUB format.

• Utilized TSUNA API for title generation, Flask for backend, and MySQL for database.

• Deployed on the Sakura server by using Docker & Nginx with CI/CD pipeline, enabling quick updates. SIDE PROJECTS

Disaster & Emergency Response - Github

• Optimized real-time disaster message classi cation with a 95% accuracy rate on testing data by training with 25+ real-life disaster messages and utilizing NLP, ML pipelines with Grid Search and Random Forest.

• Utilized ETL (Extract, Transform, Load) processes for data preparation, SQLite3 for database management, and Flask for Back-End development.

2022 NYC Airbnb Analytics - Github - Blog

• Analyzed and visualized real-world data from Inside Airbnb using Pandas, Matplotlib, Seaborn, Plotly, Folium to uncover key insights into the New York City Airbnb market.

• Used ndings to help businesses identify high-growth areas and establish competitive pricing strategies while providing tourists with top-rated listings and a ordable neighborhoods recommendations. Family Tree Web Application - Github

• Built a full-stack web application that enables users to create and share family trees by using Python, Flask to implement a REST API and MariaDB to store user information and tree designs.

• Applied Google Authentication and OAuth2 for user login and Docker-compose for deployment on a Centos server. EDUCATION

State University of New York at Oswego Oswego, NY

Bachelor's degree in Computer Science (GPA: 3.81) Jan 2019 - Dec 2022 Certi cation: IBM Data Science Specialization

TECHNICAL SKILLS

Languages: Python (primary), Go, Java, C.

Database: Mysql, MariaDB, SQLite, DynamoDB.

Tools: Git, Docker, Postman. Experienced with AWS (EC2, ECR, ECS, Codebuild, CodePipeline). Frameworks: Flask, RASA.

Front-End: HTML, CSS, JavaScript. Experienced with Bootstrap. Other knowledge: Regression/Classi cation algorithms, Computer Vision (CNN), Data Analytics/Visualization, ETL, NLP and ML pipelines.

Contact this candidate