Tanupat Phasithjirakul
Data Scientist & Software Development Engineer
Experienced in C++, Python, DevOps and Data Analysis, adept at analytical problem-solving. Fast learner, team-oriented professional with expertise in developing data pipelines and web applications.
Skills & Expertise
Key Projects
Test Cost Reduction ML Pipeline
Machine Learning • ETL • MLOps
Configured Airflow to process raw data from Redshift for training new models. Integrated MLflow to log and track model performance and metadata. Deployed trained models for production use.
LAB PC Performance Monitoring
Full Stack • DevOps • Monitoring
Monitor the test application CPU and Memory usage during test with comprehensive dashboard and analytics capabilities.
Material Verification Web App
Full Stack • Process Automation
Web application to streamline the submission, review, and approval of material verification results, generate summarized material reports for improved efficiency.
ETL Pipeline for Test Time Reduction
Data Engineering • Analytics
Coded an ETL pipeline to use historical test time data to reduce the test time in production, improving efficiency and reducing costs.
Experience
Scientist 3, Data Science
Leading data science initiatives, developing machine learning models, and implementing ETL pipelines to optimize test processes and reduce costs.
Senior Engineer, Software Development Engineering (Apps)
Coded ETL pipelines using historical test time data to reduce production test time. Developed monitoring solutions and web applications for material verification.
Engineer, Software Development Engineering (Apps)
Supported test code for hard disk drive using C++. Developed applications to improve team efficiency (PyQt). Created websites to support statistical analysis requests (Python, SPSS, JS, HTML).
Education
BE, Avionics Engineering
GPA: 2.75
Studied electronics design, airplane systems and structures
Awards
IEEE Conference Presentation
Design of a dual-band Verre de Champagne Fractal CPW antenna for LTE and aircraft altimeter application
Homelab & Technical Hobbies
🧠 Homelab Overview
My homelab is a self-hosted environment designed for AI/ML experimentation, automation, monitoring, privacy, and media streaming, all orchestrated in a well-integrated ecosystem.
🔧 Core Services
| Service | Purpose |
|---|---|
| MLflow | Tracks and manages machine learning experiments and model versions |
| Ollama | Local LLM hosting and inference server for running models like Phi-4 |
| OpenWebUI | Web interface for interacting with LLMs (Ollama integration) |
| n8n | No-code automation workflows (similar to Zapier) |
| Uptime Kuma | Self-hosted monitoring tool for service uptime and health |
| InfluxDB | Time-series database for metrics collection |
| Grafana | Visualizes metrics from InfluxDB and other sources |
| Traefik | Dynamic reverse proxy and load balancer for containerized services |
| Privoxy | Filtering web proxy for privacy and ad-blocking |
| Pi-hole | Network-wide ad-blocking and DNS filtering |
| Jellyfin | Media server for movies, TV shows, and music |
| Jellyseer | Overseer-like web UI for managing media requests to Jellyfin |
📦 Environment & Deployment
- Running via Docker and Kubernetes (k3s) cluster
- Managed by FluxCD for GitOps-based deployment
- Services exposed via Traefik with SSO authentication
- Persistent storage via TrueNAS and RAID-Z
- Automation and monitoring tightly coupled (n8n, Uptime Kuma, InfluxDB, Grafana)