Ayushi Garg
Computer Science and Mathematics student at the University of Bath with experience in programming, machine learning, and tutoring, with strong problem-solving skills and a practical approach to applying technical concepts.
Work history
Computer Science Tutor
1 Year 7 Months● Clarified key Computer Science and coding principles to enhance student understanding.
● Guided students through past papers to refine exam technique and boost confidence.
Technology Work Experience
● Shadowed software engineers and observed key processes within the Quality Assurance (QA) team.
● Attended team meetings to understand how project plans are structured and delivered.
● Gained exposure to tools and systems including APIs, LYNX, and PCAS.
● Learned how machine learning models are applied to detect fraudulent transactions and improve security.
Skills
Education
● A*A*A*A : Mathematics | Further Mathematics | Computer Science | Physics
● GCSE : 999999987 (9 in Maths)
● Additional Maths Qualification : A
Bachelor of Science - Computer Science and Mathematics
Projects
Clinical Simulation Assessment Platform | Python, NLP, SQL, SciPy
- Developed an automated web app using NLP sentence embeddings to evaluate medical student transcripts against clinical checklists.
- Implemented statistical analysis (one-sample t-test) to calculate weighted scores and automatically flag at-risk students (p < 0.05).
Gene-Drug Interaction Analyzer | Python, Data Analysis, NetworkX
- Coded a network mapping tool that analyzes genetic data against cancer drugs to identify patient risks for adverse reactions.
- Developed a weighted graph model using scientific literature volume to isolate high-risk correlations, such as CYP2C9 and Warfarin.
Heart Disease Prediction Classifier | Python, Scikit-Learn, TensorFlow/Keras
- Evaluated multiple classifiers (KNN, Logistic Regression, SVC) using confusion matrices to predict heart disease, determining SVC yielded the highest initial baseline accuracy.
- Developed a neural network using tuned ReLU/Sigmoid activation functions and addressed false-positive class imbalances to optimize overall validation accuracy and model reliability.
Affiliations
- Karate : black belt