I'm Fathan Fardian Sanum, an Informatics undergraduate student at Telkom University, Bandung with a GPA of 3.97/4.00. My primary interest lies at the intersection of Artificial Intelligence and practical software engineering. I have hands-on experience in machine learning, NLP, and computer vision, and I enjoy transforming raw data into meaningful insights and building intelligent systems.
I actively serve as a Practicum Assistant for AI, OOP (Java), and Programming Algorithm (Go) courses at the Informatics Lab. Beyond teaching, I develop projects in object detection, sentiment analysis, and I build web applications with clean architecture and intuitive UI/UX.
- Computer Vision - Real-time object detection and activity recognition using YOLO, RF-DETR, and MediaPipe
- Natural Language Processing - Sentiment analysis and text classification using BERT, IndoBERT, TF-IDF, and word embeddings
- Deep Learning - CNN, LSTM, transformer-based architectures with hyperparameter tuning
- Multimodal AI - Integrating vision, TTS, and LLM APIs for inclusive and accessible learning systems
- AI Ethics Education - Community engagement in promoting responsible and ethical AI usage
| Project | Description | Stack |
|---|---|---|
| BISIK | Inclusive multimodal education platform with Gemini Vision API, Unsplash, and Google TTS for low-resource learners | Next.js, Gemini API |
| Sentiment Analysis on TikTok | Benchmarked 8 NLP models on 6,853 Indonesian comments; best: IndoBERT Acc 86.43%, F1 0.86 (GEMASTIK 2025) | Python, HuggingFace |
| AI Surveillance System | Real-time face detection, activity recognition & anomaly detection module | Python, YOLO, MediaPipe |
| PPE Compliance Monitoring | RF-DETR trained on SH17 (17 classes), outperformed YOLOv8 with mAP@50: 0.580, Macro F1: 0.606 | Python, RF-DETR |