Muhammad Kemal Fasya

I’m a fresh graduate in Informatics with a strong interest in machine learning and data-driven problem solving. Through my projects and coursework in data mining and machine learning, I’ve built a solid foundation in developing models and communicating ideas clearly. I work well under pressure, take initiative to lead workgroups, and embrace kaizen—learning from challenges and mistakes.

I focus on practical, end-to-end modeling, especially for computer vision and forecasting. I’ve implemented an OCR/digit recognition approach using YOLOv8 for region-of-interest detection with LSTM + CTC Loss, and I developed an automatic plate number recognition pipeline using YOLOv8 with real-world dataset preparation and evaluation. I also built a time-series forecasting model to automate bandwidth allocation for internet voucher sales.

My experience includes an internship as a Machine Learning Engineer where I used Teachable Machine for introductory learning and worked with YOLOv8 and OCR in object detection and object recognition. I also completed an intensive applied machine learning and data science cohort, strengthening my Python-based model development and teamwork. I was selected as the Best Speaker during a capstone final presentation at Bangkit Academy.