Bruce Power

Data Science Intern

• Developed front-end applications using Dash and Streamlit to create interactive dashboards, enabling real-time data visualization and driving key insights for stakeholders, resulting in a 20% increase in decision-making.
• Leveraged Generative AI frameworks such as LangChain and LlamaIndex to build innovative AI applications, leading to a 25% increase in the generation of automated responses and solutions.

Massachusettes Institute of Technology

Data Engineering Intern

• Developed web apps with real-time data integration using JavaScript, Node.js, and REST APIs, providing users with accurate and up-to-date visual insights, improving user engagement by 30%.
• Designed and implemented user-centric visualizations by applying principles of typography, chart forms, and color theory, resulting in a 40% increase in user satisfaction as measured by survey feedback

Dartmouth College

Data Engineering and Business Intelligence Intern

• Achieved high data processing efficiency, as measured by reducing ETL processing time by 30%, by designing and developing batch and real-time data pipelines using Java and Apache Spark
• Facilitated a seamless transition to cloud infrastructure, as measured by a 50% reduction in on-premises infrastructure costs, by shifting data storage and processing to GCP tools such as BigQuery and Dataflow

University of Toronto

Machine Learning Intern

• Developed neural network models to analyze and predict sleep patterns using data from wearable devices and mobile health applications, resulting in a 25% improvement in sleep detection accuracy.
• Conducted a comparative analysis of SVM, Random Forest, and CNN for sleep stage classification, achieving a 92% accuracy with CNN. This resulted in a 15% performance improvement over the baseline
• Created Tableau and Power BI dashboards, improving decision-making and reducing reporting time by 20%

University of Western Ontario

Data Science Intern

• Developed a Generative Adversarial Networks model using Tensorflow and PyTorch to generate detailed mages of MRI Scans gathered from experimental data with an accuracy rate of 85%.
• Optimized the performance of a multi-terabyte research database, which reduces the data retrieval time by approximately 40%, by administering frequent checks and updates on SQL and NoSQL systems.