Shibaura Institute of Technology, Tokyo, Japan

Master’s Course, Electrical Engineering and Computer Science

Data Science/Engineering Laboratory – NLP Researcher

Current GPA 3.9 / 4

Aug 2019 – Sept 2023

Institut Teknologi Sepuluh Nopember, Surabaya, Indonesia

Bachelor of Engineering in Computer Engineering

GPA 3.75 / 4

Jul 2025 – Aug 2025

Full-Stack Developer – ocrlogbook.com↗

Joint development with an Air Hong Kong senior captain to build an automated OCR-based flight-log extraction tool.

• Partnered with an Air Hong Kong senior captain who needed a faster, more accurate way to record flight logs.
• Designed and deployed an automated system to extract and manage structured flight data from image uploads.
• Built a serverless OCR pipeline using AWS S3, Lambda, API Gateway, CloudFront, and Route 53, integrating Amazon Textract and developed a responsive HTML/CSS/JavaScript frontend.
• Reduced manual entry time from minutes to seconds per record, enabling batch processing and full end-to-end automation on ocrlogbook.com.

Apr 2024 – Aug 2025

Research Publication – Enhancing Automated Essay Scoring Model’s Interpretability↗

Published on the 6th Joint International Conference on AI, Big Data and Blockchain (AIBB 2025)

• Proposed a rubric-aligned automated essay scoring method and model, leveraging triplet loss with teacher-defined rubric as anchors to improve interpretability.
• Introduced Dual-Anchor Triplet Loss function, enhancing score separation and training efficiency.
• Built and trained models on top of DeBERTa, on a dataset containing large numbers of student essays.
• Achieved performance comparable to state-of-the-art models, outperformed larger LLM approaches, and published at AIBB 2025, Istanbul.

Nov 2025

AWS Serverless Website with CI/CD – josephsaido.com↗

Personal portfolio and resume built on a fully serverless AWS architecture with a custom frontend.

• Built a secure, scalable, and automated website using AWS serverless services and CI/CD best practices.
• Designed a serverless architecture using S3, CloudFront, Route 53, and ACM for global delivery.
• Implemented serverless backend processes with Lambda and DynamoDB.
• Set up multi-environment CI/CD pipelines using GitHub and CodePipeline.
• Provisioned the entire infrastructure with Terraform.
• Developed a modern frontend with HTML, CSS, and JavaScript.
• Delivered a globally distributed portfolio with quick deployments and fast content delivery worldwide.

Jun 2025

Desktop AI Companion App – MirAILive↗

An open-source desktop AI companion that observes screen context and responds in speech.

• Designed and implemented a real-time, vision-language-driven companion that combines contextual awareness, dialogue generation, and expressive voice output.
• Built a desktop overlay app in Python (PyQt, OpenCV) integrating vision-language models from the OpenAI API and local F5-TTS for natural voice responses.
• Managed message context and structured prompting for coherent dialogue.
• Designed and illustrated the virtual character.
• Published an open-source AI companion capable of understanding on-screen context and engaging in real-time, speech-based interaction.

Sept 2022 – Aug 2023

Bachelor’s Final Research Project – Hand Pose-Based Video Game Control

A machine learning system that lets players control racing games using only webcam-detected hand gestures.

• Developed a machine learning-based game controller that uses webcam hand-pose recognition for natural racing game interaction.
• Implemented and benchmarked multiple CNN architectures (ResNet, EfficientNet), exploring trade-offs between accuracy and inference speed, achieving 98% accuracy with low-latency real-time control.
• Integrated models into a fully functional desktop application for user testing and demonstration.
• Conducted a User Experience Questionnaire (UEQ) study, showing positive feedback on usability, clarity, and enjoyment.

Apr 2022 – July 2022

Android App Developer for ‘rePlasc’ – Waste Recycling App

Bangkit Academy, a Google-led Career Readiness Program, Capstone Project

• Prototyped and co-developed a mobile application as part of a multidisciplinary capstone team.
• Implemented key features including RESTful API integration and responsive UI components.
• Integrated a TensorFlow Lite model for real-time image classification directly on the device.
• Collaborated closely with teammates to ensure system functionality and deliver the final product to a partner company.