Curating weekly, monthly, and yearly financial reports for franchisee on ExcelReconciling 100+ sales reports of customer transactions daily on OracleConstructing weekly employee schedules and organizing office materials

Developed a neural network model from scratch (no TensorFlow or PyTorch) using NumPy to classifyhandwritten digits from the MNIST dataset, achieving 85% accuracy.• Preprocessed and normalized dataset of 60,000 training images and 10,000 testing images with pandas, ensuringefficient data input and optimal model performance.• Implemented ReLU and softmax activation functions to introduce non-linearity and generate output probabilitydistributions, enhancing model decision-making.• Fine-tuned hyper-parameters, including learning rate and iterations, to optimize weight and bias updates, boostingmodel training efficiency.

Developed several infographics and flyers to welcome interaction on social media for non-profitorganization,Collaborated with other members to edit and draft website in HTML, CSS, and JavaScript,Contributed to the simplification of over 15 research articles