Building AI systems that are useful, testable, and real.
I am a Computer Engineering student at IOE, Tribhuvan University, working across machine learning, computer vision, NLP, and data science. I enjoy turning research ideas into working systems and learning how those systems perform in real conditions.
I'm Aaditya Kumar Singh, a Computer Engineering student at IOE, Tribhuvan University focused on Artificial Intelligence, Machine Learning, and software engineering. My work ranges from computer vision systems and data analysis to retrieval-augmented generation and Nepali language model research.
I care about understanding why a model works, how it should be evaluated, and how it can become a usable application. I am currently exploring RAG-based root-cause analysis, NLP, Linux, networking, cloud infrastructure, and DevOps. I also serve as Chair of Club Administration at the Rotaract Club of Kathmandu.
RAG-Based Fault Diagnosis Assistant
Research project comparing vector and vectorless retrieval for automated root-cause diagnosis in microservice systems using RCAEval.
Real-Time Fruit Freshness Detection
Two-stage computer vision system using YOLOv8 for detection and MobileNetV3 for fresh versus rotten classification.
Tisa: Nepali Generative Language Model
Research involvement with a 116M-parameter Nepali decoder-only language model, focusing on architecture and downstream evaluation.
Breast Cancer Prediction System
End-to-end Random Forest prediction system with preprocessing, model evaluation, FastAPI backend, and React frontend.
AES-Based Image Steganography
Secure communication tool combining AES encryption with LSB steganography to hide encrypted messages inside images.