Professional Experience
Aitomatic, Inc.,
Remote (Vietnam)
Software Product Engineer
Mar 2021 - Present
- Accomplished the development of a question-answering product by utilizing Large Language Models (LLM), leading to enhanced document interpretation and agent reasoning processes.
- Achieved cost-effective scalability in Kubernetes clusters by employing Terraform and Terragrunt, leading to automatic utilization of the most suitable spot instances based on cluster usage.
- Achieved robust and reliable CI/CD pipeline by utilizing Github Actions, leading to significantly improved uptime and deployment time.
- Facilitated UI-based ML model building and data inference by developing a service API with FastAPI.
- Supported product development by constructing framework libraries based on OOP philosophy.
- Enhanced anomaly detection by developing a weekly/daily ETL pipeline that ingests and transforms coldchain data from customers into parquet format, leading to positive customer feedback.
Python
AWS
Kubernetes
Terraform
FastAPI
LLM
Github Actions
SP Digital,
Singapore
Senior Data Engineer
Feb 2019 - Feb 2021
- Successfully upgraded data infrastructure to a newer version without downtime by carefully managing and scheduling updates over 3 months.
- Decreased model training time by 30% and increased accuracy by 45% by creating and managing a system that gathers, transforms, and feeds useful information to a predictive model for electric meter replacement.
- Improved code performance and readability by establishing SQL coding standards and recommending an effective platform for SQL code management.
- Reduced the Mean Time To Repair (MTTR) of data lakes by developing Python scripts for automatic data monitoring.
- Contributed to the creation of an 80% accurate 48-hour price prediction model by developing hourly Kubernetes jobs that ingest electric price data.
Python
Java
Talend
Hadoop
Spark
Azure
Kubernetes
Honestbee,
Singapore
Data Engineer
Sep 2016 - Jan 2019
- Ensured clean and accurate reporting for the BI team by building a data pipeline that processes and stores user, order, behavior, and third-party data.
- Increased overall performance by 25% for product ranking and recommendation systems by optimizing queries.
- Ensured uninterrupted data services by maintaining the high availability of legacy data infrastructure.
Python
AWS
Airflow
Postgres
Docker
Mesos
Looker
Review: Nick Young, Head of Data
Education
Master degree in Knowledge Engineering, NUS, Singapore [2015 - 2016]
Bachelor degree in Software Engineering, FPT University, Vietnam [2009 - 2013]
