Software Engineer (AI/ML)
Machine Learning Engineer (contract)
Achieved 36% reduction in maintenance costs by leading the development and implementation of machine learning models to forecast machine incidents. Built predictive models using Python and XGBoost, leveraging historical incident data, hardware logs, and customer usage patterns to optimize system reliability and operational efficiency.
Machine Learning Engineer (full-time) · Band L1
Increased revenue by 15% through development and support of a PySpark application for analyzing KBank customer statements. Led stakeholder collaboration to present data insights, identify requirement gaps, and recommend strategic solutions.
Innovation Engineer (Internal Re-organization) · Band L1
Achieved 20% productivity improvement across the research team by developing a React-based internal web application that streamlined AI experimentation workflows. Enabled research engineers to efficiently experiment with prompts and multiple LLM models, reducing model testing time and eliminating manual configuration overhead.
AI/ML Engineer (Internal Re-organization) · Band L1
Achieved 40% faster model deployment cycles by leading the productionization of research models from data scientists into scalable production systems. Developed comprehensive MLOps pipelines including code quality check, automated unit test generation, and automated metadata generation. Collaborated closely with data science teams to bridge the gap between experimental models and enterprise-ready AI solutions, ensuring robust performance monitoring and seamless integration with existing banking infrastructure.
Bachelor of Science, Computer Science · GPA 2.78/4.0
Python, Java, C++, Rust, JavaScript, TypeScript, React.js, Next.js, Node.js, Express.js, Flask, Django
PySpark, Databricks, Bash, SQL
Pandas, TensorFlow, PyTorch, XGBoost, MLFlow, RAG, LangChain, LangGraph, Gemini, OpenAI, Anthropic, Deep Research
Git, Docker, GCP, Azure, Databricks, GitHub Action
Thai (Native), English (Intermediate)
Food waste is a significant global issue, with many households discarding ingredients due to lack of recipe ideas. This project solves this problem by using generative AI to identify ingredients from images and create personalized recipes based on what's available.
Keywords: Google Gemini, Multimodal, Structured Output, Prompt Engineering —Kaggle