Ai Engineer Information Technology

TRUNG LE MINH

As an individual committed to responsibility and perpetuat

leaming. I take pride in my ability to adapt flexibly and continuously acquire knowledge to meet the diverse requirements of each project. My enthusiasm and

willingness to embrace change are vital qualities that enable me to consistently achieve high work eficiency.

Gender: Male

Date of birth: 20/03/1999

Email: ***************@*****.***

Phone: 098*******

Address: Hue, Viet Nam

OBJECTIVE

EDUCATION

Information technology

FPT Greenwich

University

07/2017-01/2022

WORK EXPERIENCE

AI Engineer

FPT Software 10/2023-Present

LLM Chatbot Development Project:

Developed an AI chatbot to answer customer queries about products, warranty policies, and gather user trend data. Built and maintained the backend using FastAPI, MongoDB, and Docker, ensuring high scalability and reliability. Collaborated with cross-functional teams to design RESTful APIs and database schemas optimized for

performance.Leveraged advanced LLMs (Llama 3, Gemini, and GPT-4) to enhance conversational capabilities and accuracy. Analyzed system bottlenecks and optimized query

performance, leading to a 30% improvement in response time.Created unit and integration tests using PyTest to maintain code quality and ensure robust functionality. Actively participated in Agile processes, including daily stand- ups, sprint planning, and retrospectives.

Technologies Used: Python, FastAPI, MongoDB, Docker, Llama 3, Gemini, GPT-4, PyTest.

SKILLS

Programming Languages

Python, Java, C#, HTML & CSS, Javascript,

ReactJS, Flutter, Git

Databases etc

MS SQL Server, PostgreSQL, MongoDB,

Firebase

Programming Tools

Visual Studio Code, Eclipse, Pycharm,

Android Studio, Docker

Other

Work with GPU in AWS

AI Engineer

FPT Software 2/2024-10/2024

Website For Visualize Data Project:

Developed a website using Dash for interactive data visualization and analysis. The platform allows users to explore and visualize data through a variety of charts and graphs, making complex data more accessible. Integrated machine learning models, particularly LSTM (Long Short-Term Memory), for predicting future trends based on historical data. The project includes real-time data updates, enabling dynamic interaction and exploration of forecasted results. Key Features: Interactive data visualization with customizable charts. LSTM-based prediction for future data trends. Real-time data updates with intuitive visual analysis. User-friendly web interface built with Dash.

Technologies used: Dash, LSTM, Plotly, Pandas, NumPy, Scikit- learn.

AI Engineer

FPT Software 10/2023-2/2024

Document-Based Chatbot Project:

Designed and implemented a backend system for a chatbot that answers user queries based on uploaded documents. Developed RESTful APIs using FastAPI, ensuring secure and seamless integration with frontend applications.

Integrated LLMs (Llama 3, Gemini, GPT-4) to enhance the chatbot's contextual understanding and response

generation.Designed ERD and optimized database queries, reducing latency and improving data retrieval efficiency by 20%.

Deployed and managed containerized applications with Docker, adhering to best DevOps practices.Wrote and executed comprehensive unit tests, ensuring 95% code coverage and minimal system downtimes.

Collaborated closely with stakeholders to align technical development with business goals and user needs.

Technologies Used: Python, FastAPI, MongoDB, Docker, Llama 3, Gemini, GPT-4, PyTest.

AI Engineer

FPT Software 4/2023-10/2023

Handwriting Generation Project:

Developed a handwriting generation system capable of creating text in various user-selected styles using GAN models.

Conducted comprehensive data preprocessing to ensure high- quality inputs for model training, including noise reduction and normalization using OpenCV.

Optimized the GAN architecture to generate realistic and style- consistent handwriting outputs.

Evaluated model performance using metrics such as SSIM

(Structural Similarity Index Measure) and ensured smooth integration into downstream applications.

Technologies Used: Python, GANs (e.g., DCGAN, CycleGAN), OpenCV.

AI Engineer

FPT Software 10/2022-4/2023

OCR Project (Team lead):

Designed and implemented a system to detect and recognize Japanese characters in images with high accuracy.

Preprocessed image data by enhancing contrast, removing noise, and optimizing for character segmentation.

Developed algorithms to evaluate model accuracy, leveraging precision, recall, and F1-score metrics for robust performance monitoring.Integrated state-of-the-art tools like MangaOCR, EasyOCR, and YoloX for precise character detection and recognition, improving detection efficiency by 30%. Fine-tuned and tested the OCR system to handle various image qualities and character complexities.

Technologies Used: Python, MangaOCR, EasyOCR, YoloX. AI Engineer

FPT Software 3/2022-10/2022

Face Detection and Tracking Project:

Implemented a system for real-time face detection and tracking using Yolov7 for identifying and tracking human faces in video streams. Integrated Dlib for precise face pose estimation to determine the orientation and angle of the detected faces.

Developed a web socket interface to stream real-time video with tracking information overlay. Used PostgreSQL to store tracking data, allowing for efficient querying and retrieval of face tracking history and analytics.

Key Features: Face detection and tracking using Yolov7. Pose estimation via Dlib to identify face orientation. Real-time video streaming through web sockets. Storage of face tracking information in PostgreSQL for later analysis.

Technologies used: Yolov7, Dlib, PostgreSQL, WebSocket, Flask, OpenCV.

AI Engineer

FPT Software 6/2021-3/2022

Computer Vision Project:

Implemented detection and classification of objects with unclear characteristics Applied varicus image preprocessing methods for data training and detection preparation.

Customized parameters to achieve the highest possible model accuracy.

Developed and utilized algorithms to calculate model accuracy for both single label and multi-label cases.

Technologies used: Yolov7, YoloX, PatchCore, Resnetgo, and VGG16.

Java Developer

FPT Software 2/2021-6/2021

Engaged in a web maintenance project utilizing Java Spring and ReactJS.

Assumed the role of a programmer executing tasks as per client requirements.

Applied efficient development strategies to meet project objectives.

Outcome: Successfully completed key modules, ensuring quality and adherence to deadlines.

Final Project In University

SmartHome Mobile

App (IoT)

8/2020-12/2020

Create a mobile application that can control devices base on the NodeMCU ESP8266. Main function: diplay the status of the devices (open or close, turn on or turn off), display information, change the operating status of the devices.

- Technology use:

+ Android Studio

+ Firebase of Google service

+ Java, C/C++

Flutter (Freelancer)

Mobile App

Participated in the development of mobile applications using Flutter, including: A driver tracking application for a fruit shop’s delivery service (integrated Google Maps API to display real- time driver locations), A voice translation and OCR-based text recognition app (translating and processing text from images). Built and managed the backend for applications, ensuring stable and efficient system operations.

Designed, tested, and deployed AI models, integrating TensorFlow Lite into applications to optimize AI performance on mobile platforms.Utilized Firebase to build and manage databases, supporting real-time data synchronization and other cloud features.Collaborated with the team to test, optimize, and enhance application performance before deployment. Technology use: Flutter, Firebase Realtime Database, Firebase Cloud Firestore, Google Maps API, Firebase APIs, TensorFlow Lite.

Contact this candidate