Skip to content
~/hth

~/experience

Experience

Where I've worked, what I did there, and the skills I use.

DevOps Engineer @ PMAX - Total Marketing Partner (opens in a new tab)

Sep 2026 – Present·3 mos

CI/CD, self-hosted runners, servers and monitoring for PMAX's four internal platforms.

  • Set up and run self-hosted GitLab Runners (Docker executor) on Linux VPS and AWS Lightsail, so our pipelines have their own build capacity, Docker layer caching and no shared-runner minute limits.
  • Built and maintain the GitLab CI/CD pipelines (build image, push to AWS ECR, deploy) for staging services on CreativeOS, MediaOS, CompanyOS and FinOS. Deploys now run on merge instead of by hand, and a release takes about 5 minutes instead of 20.
  • Manage DNS records, TLS and proxy settings on Cloudflare, and script routine server and deployment tasks in Bash.
  • Built a Google Chat bot for the company workspace with Google Cloud APIs. Only users with the right role can write to our AWS PostgreSQL database, and every interaction (user, action, status) is logged to Google Sheets. It runs on AWS Lightsail.
  • Built Grafana dashboards for Amazon RDS PostgreSQL from CloudWatch metrics and logs (CPU, connections, storage, latency, slow queries), and tested them on staging before rolling them out to production.
  • Wrote runbooks and incident notes in English for deployment, rollback and Runner recovery.
  • GitLab CI/CD
  • GitLab Runner
  • Docker
  • AWS Lightsail
  • AWS ECR
  • Cloudflare
  • Bash
  • Grafana
  • CloudWatch
  • Amazon RDS
  • Google Cloud APIs

Analytics Engineer Intern @ GiaoHangNhanh (GHN) (opens in a new tab)

Mar 2026 – Jun 2026·4 mos

Internship · Ho Chi Minh City

Data pipelines, SQL analysis and reporting for logistics operations.

  • Wrote BigQuery SQL (multi-table joins, CTEs, window functions) for ad-hoc analysis, and automated exports to Google Sheets with Python. Reports that took 3–4 hours to put together by hand now take a fraction of that.
  • Built and maintained daily ETL/ELT pipelines with Apache Airflow in Docker, turning raw logistics data into dimensional tables for analytics.
  • Built AppSheet apps to control who can see which reports, and Looker Studio dashboards (with pivot summaries) that several departments use to track logistics KPIs instead of reporting by hand.
  • Apache Airflow
  • Docker
  • BigQuery
  • SQL
  • Python
  • Looker Studio
  • AppSheet
  • Google Sheets

Data Processing Intern @ FPT Information System HCMC (opens in a new tab)

Oct 2025 – Mar 2026·6 mos

Internship · Ho Chi Minh City

Backend data work for the HCMC Metro and government websites.

  • Processed and organized backend data for the HCMC Metro System and government websites during the provincial merger, using SQL joins and aggregations over OLTP tables to produce clean, query-ready datasets.
  • Helped build a Metro passenger analytics system: aggregated passenger counts per station in 30-minute slots from 5:00 AM, and made pivot reports to compare stations and times of day.
  • Helped build statistical and predictive models for passenger volume, and checked their output against historical ridership.
  • Built a web app where stakeholders can compare forecast and actual passenger numbers on their own.
  • SQL
  • OLTP
  • Data processing
  • Predictive modeling
  • Web app

skills

Skills

Languages & Scripting
PythonJavaSQLBashShellYAMLJSON
Linux & Systems
Linux administrationSSH keysFirewall rulesMicrosoft Windows
Networking
TCP/IPDNSSubnettingAWS VPC & subnetsCloudflare (DNS, TLS, proxy)
Cloud & Monitoring
AWS (EC2, Lightsail, S3, IAM, VPC, Lambda, RDS, ECR, CloudWatch)GrafanaGoogle Cloud (familiar)Azure (familiar)
Containers & CI/CD
DockerDocker ComposeGitLab CI/CDGitLab Runner (Docker executor)GitHub ActionsGitOps
Data
Apache AirflowPostgreSQLMySQLSQL ServerBigQueryMinIO
BI & Reporting
Looker StudioPower BI (DAX)AppSheetGoogle Sheets
Tools
GitJiraClaude CodeOpenAI CodexAI-assisted tools