About

I am passionate about development, having turned my casual interest into an exciting journey. While I might not see myself as a creative genius, I find joy in transforming ideas into fun, interactive experiences.

I am currently working towards a Master of Engineering in Data-Intensive Intelligent Software Systems, fully immersing myself in the realm of data. My primary goal is to capture raw data from its sources, refine it, and transform it into impactful and practical insights.

I strongly believe in the power of teamwork and learning from different perspectives. Collaboration is where we find the best solutions, and I prioritize open communication and adaptability in my work.

When I’m AFK, you’ll find me immersed in the rhythms of life—my favorite tunes are my sanctuary. Whether I'm coding, analyzing data, or simply being lost in a melody, I embrace new challenges with enthusiasm. Let's connect and craft something extraordinary together!

Experiences

  1. 2023 - Present

    Create a streamlined planting and harvest management system to enhance farm operations and align crop output with market demand. Design automated workflows for key tasks like planting, watering, and harvesting across franchise locations to ensure consistency and reduce manual oversight. Develop a dashboard for real-time visibility into crop status, task progress, and inventory, enabling data-driven decision-making. Promote sustainable practices and optimize resource use throughout farm management.

    • Nuxt.js
    • TypeScript
    • Express.js
    • Vuetify
  2. June - Aug 2024

    Created data pipelines to process raw sensor data from nursing home devices, providing insights into resident health. Analyzed vital signs and movement data to establish health metrics and monitor time spent in different locations. Developed APIs and an interactive dashboard for caregivers to track 24-hour health data and observe trends over time.

    • Python
    • Numpy
    • Pandas
    • Matplotlib
    • Flask
    • React.js
  3. 2022 - 2023

    Conducted data collection in rural Lao PDR, gathering vital health information to support research and public health initiatives. Prepared and standardized large text datasets from local dialects, ensuring accuracy for analysis. Analyzed findings and created reports, providing critical insights to the Ministry of Health and Home Affairs for informed policy-making and health interventions.

    • Python
    • Excel
    • Numpy
    • Matplotlib
    • SurveyCTO
  4. Sept - Nov 2021

    Developed scalable and maintainable mobile applications using Flutter and CLEAN Architecture, achieving 80% code coverage through unit testing. Contributed to upgrading the Trakref app from Flutter 1.x.x to 3.x.x, improving performance and user experience. Created an automated testing application with Appium to boost testing efficiency. Gained expertise in Agile and SCRUM methodologies, enhancing team collaboration and project management.

    • Flutter
    • CLEAN Architecture
    • Appium
    • Insomia

Projects

  1. Oct 2024

    A first version of my portfolio site built with Nuxt.js.

    • TypeScript
    • Nuxt.js
    • TailwindCSS
    • Vercel
  2. Feb 2024

    Leveraging CUDA for GPU acceleration, this project enhances galaxy data analysis by efficiently computing histograms of galactic coordinates. It processes large datasets to capture spatial distributions of real and simulated galaxies, minimizing global memory access by employing shared memory within each CUDA block. Benchmarks demonstrate substantial speedups, achieving a 2-second runtime on a high-CPU setup (98% utilization) and 0.78 seconds on an NVIDIA Tesla V100. This approach enables fast, accurate galaxy data processing, making it ideal for high-performance computational tasks.

    • CUDA C
    • NVIDIA GPU
    • Data Parallelism
  3. Feb 2024

    Focusing on efficient deployment on the Arduino Nano 33 BLE Sense, this project builds an image classification model optimized through 8-bit quantization for resource-limited hardware. Real-time images from a camera sensor are preprocessed by resizing, normalizing, and transforming them into a 784-dimensional input vector. The model is converted to an 8-bit format using TensorFlow Lite's post-training quantization, significantly reducing memory usage while maintaining high classification accuracy. This approach enables efficient and responsive image classification within the Arduino Nano’s memory constraints.

    • Python
    • TensorFlow Lite
    • Arduino
    • 8-bit Quantization
  4. Feb 2024

    Clustering human activity data from accelerometer and gyroscope readings, this project applies K-Means and DBSCAN methods, enhanced through dimensionality reduction techniques like PCA and t-SNE, to achieve precise activity grouping. The approach demonstrates how tuning and dimensionality reduction impact cluster structures, providing valuable insights into patterns of activity recognition.

    • Python
    • Pandas
    • Scikit Learn
    • K-Means
    • DBSCAN
    • PCA
    • t-SNE
  5. Feb 2024

    This was created to help schedule sauna sessions at our student dormitory (Tys). Our sauna is spacious enough for individual use, so I thought it would be great to share it among us international students. In the past, we often forgot our booking times or who had reserved the sauna, leading to some confusion. To avoid overlapping bookings and missed turns—and to add a bit of fun with a jokes command for entertainment—this system will help us track reservations easily through a bot command.

    • JavaScript
    • Express.js
    • Axios
    • WhatsApp-Web.js
    • Google App Script
    • AWS
  6. Sept 2023

    Collaborated with the Neonatal Intensive Care Unit at Turku University Hospital, focusing on extracting and processing Edi signals from medical devices. This involved transforming signals into the frequency domain and developing predictive models to forecast apnea episodes up to one minute in advance. The project emphasizes continuous algorithm optimization, with ongoing efforts to document findings in a research paper.

    • Python
    • Numpy
    • Pandas
    • Matplotlib
    • Seaborn
    • Scikit Learn
    • Keras
  7. July 2022

    A transparent and genuine charity web application uses blockchain technology and smart contracts to ensure secure, decentralized cryptocurrency donations on the Ethereum Test Network. The application enhances transparency by enabling donors to track transactions directly, ensuring each charity project’s funds are managed openly without reliance on a central authority.

    • Vue.js
    • Nuxt.js
    • Vuetify
    • Express.js
    • GraphQL
    • Apollo
    • MongoDB
    • Solidity
    • Ethereum
    • MetaMask