My post announcing that I joined ASX was dated 2026-11-07 10:32 UTC.
I wrote, “I joined ASX today.”
I shared my intention: “Looking forward to meeting folks on this platform to make meaningful connections.”
My post does not specify what “ASX” refers to, and it gives no separate date for when I joined.
I am Bhishma Bhandari, based in Bhimdatta, Far-Western Development Region, Nepal. My headline is “Catering My Interests.” No current or past positions are listed on my profile.
I am pursuing a Bachelor of Engineering - BE in Computer Engineering at Kathmandu University (KU): Bachelor of Engineering - BE, Computer Engineering (Jan 2022 to Jan 2026).
I studied +2 Science, Medicine at Kathmandu Model Secondary School: +2 Science, Medicine (2020 to 2022).
- My listed skills include Convolutional Neural Networks (CNN), Computer Vision, Data Annotation, Dataset Preparation, Exploratory Data Analysis, Pandas, Data Visualization, Supervised Learning, Deep Neural Networks (DNN), Amazon Web Services (AWS), Cloud Applications, Feature Engineering, Object-Oriented Programming (OOP), Python (Programming Language), and Event Management.
- I earned the Feature Engineering for NLP in Python certification: Feature Engineering for NLP in Python, Nov 2025, DataCamp.
- I earned the Object Oriented Programming with Python certification: Object Oriented Programming with Python, Nov 2025, DataCamp.
- I earned the AWS Cloud Practitioner certification: AWS Cloud Practitioner (CLF-C02), Oct 2025, Amazon Web Services (AWS).
- I earned the Data Manipulation with Pandas certification: Data Manipulation with Pandas, Oct 2025, DataCamp.
- I earned the Supervised Learning with scikit-learn certification: Supervised Learning with scikit-learn, Sep 2025, Python.
One of my projects is EyeSpy: AI-Powered Exam Proctoring System.
For EyeSpy, I developed an intelligent object detection system using YOLOv8 and PyTorch to automate cheating detection in physical classroom exams through real-time video surveillance.
The system is intended to address the limitations of manual invigilation through accurate, automated monitoring. It performs real-time multi-object binary classification of cheating versus non-cheating behaviors, uses Albumentations data augmentation, and was deployed as a lightweight model with ONNX export for cross-platform compatibility.
My EyeSpy model result was: Trained YOLOv8s model on 880+ custom-labeled images achieving 92% precision and 85% recall.
I trained the EyeSpy model on this hardware: Trained on Kaggle's Tesla T4 GPU.
The EyeSpy technical stack is Python, PyTorch, YOLOv8 (Ultralytics), OpenCV, Albumentations, NumPy, and ONNX. Its stated impact is to reduce reliance on manual proctoring, minimize human error, and enable scalable exam monitoring in educational institutions. The project repository is https://github.com/hridayanshu236/EyeSpy.