+1-970-***-****

MATTHEW ELIOT

Do Thanh Thong

Junior AI Engineer

094*-***-*** – ad2sxv@r.postjobfree.com

Summary

I am currently employed as an AI Engineer with 1 year of experience in the implementation of Computer Vision and NLP models into software applications. I have knowledge about Machine Learning algorithm, data preprocessing techniques. I have hands-on experience in Software Development, including developing Python Web APIs, Docker deployment, working with SQL databases, Microsoft Azure.

Education & Certificate

GPA: 8.6 - Bachelor of Electrical & Electronic Engineering – Ho Chi Minh City University of Technology.

Won 4th Prize at Quy Nhon AI Hackathon 2022.

Technical skills

Deep Learning, Machine Learning

Data Preprocessing, Feature Engineering

Python API

SQL database

Docker, Linux, Azure

ETL pipeline

Project

Table Extraction OCR Image Processing, PaddleOCR, FlaskAPI. FPT Software, Oct 2023 – Present Extract handwritten number and printed Japanese PDF tables and format them into a CSV file.

• Used image processing: enhancing contrast, sharpness, brightness.

• Used image morphological operations to detect the vertical and horizontal lines of table.

• Extracted table structure, extracted cells from the table.

• Recognized handwritten number and printed Japanese from those cells.

• Implement rule-based approach to convert OCR results from cells into CSV table.

• Integrated with backend using Flask.

Chatbot FAISS model, FastAPI, Docker, Bash. FPT Software, Aug 2023 – Oct 2023 Developing an AI Chatbot Production that finds answers to user’s questions using similarity search and provides responses through a user-friendly interface.

Built an AI Model, AI database for answering questions using similarity search.

Developed basic features for AI Chatbot, including adding data, running answers, and evaluations.

Created feature to update AI model, AI database from customer’s chat history.

Implemented version control for the AI database.

Implement APIs to integrate with Backend.

Worker Compliance Checking AI model, MySQL, Triton, Redis, Flask, Docker. FPT Software, Feb 2023 – Apr 2023 Used computer vision techniques such as detection, tracking, keypoint detection and rule-based system to check whether workers violated established regulations or not. There are regulations about uniforms, masks, helmets, working posture. Issued warnings if violations were detected.

Built a Uniform service to detect uniform, helmet, mask used by worker.

Implemented a Rule-based Business service for compliance checking.

Created MySQL database to store uniform, helmet embeddings and information about violating worker.

Serving model using Triton.

Streaming data using Redis.

Implement APIs to register, update valid uniforms and helmets. Review Analytics – Top2 solution FPT AI Hackathon Pytorch, Transformer. Jun 2022 – Sep 2022 Built an AI model to classify and rate tourist’s reviews on the following aspects: entertainment, accommodation, restaurants, food, transportation, shopping.

Conducted data labeling, data cleaning, feature engineering.

Implement a baseline model using PhoBert and multilingual model.

Implemented metrics.

Using K-fold cross validation to identify incorrectly predicted and analyzed data.

Customized head of PhoBERT model.

Customized and implemented label smoothing.

Implemented pseudo-labeling.

Fine-tuned hyperparameters.