At IT Consortium, I contributed to research and development of data-driven solutions for financial technology applications. I worked on an unsupervised machine-learning approach to fraud detection using transaction data and user-level behavioral patterns.
I cleaned, preprocessed, sorted, and segmented interleaved financial transaction records. I also engineered time-based and transaction-level features and used exploratory data analysis to examine user activity, spending, and temporal patterns.
For this work, I explored feature selection and dimensionality reduction, including PCA, as well as Isolation Forest for identifying unusual transaction behavior.
In my Adolescent Reproductive Health Study, I designed and administered questionnaires, cleaned data, and conducted quantitative analysis using SPSS. I’ve also worked on an AI-powered expense tracking system to analyze spending patterns and predict future expenses.

