AI Data Analyst with a Master’s in Data Science. She has hands-on experience in enhancing model accuracy and efficiency at Domain, and developing an OCR-based receipt management system. She has also led data pipeline projects and created a counterfeit tagging system. Her notable achievements include awards in data science competitions and significant contributions to research and mentorship in the field. Her skills include Python, Microsoft Azure, AWS, and various data analytics tools.
– Engineered a split BERT model pipeline, elevating
prediction accuracy by 90%.
– Designed and implemented a caching pipeline, optimising
training speed by 90%.
– Developed a model pipeline leveraging large language models
to forecast property prices from listing descriptions,
employing PyTorch and Huggingface
– Pioneered an OCR-based receipt management system using
Microsoft Azure and Python.
– Automated client workflows and generated insightful reports
utilising Python and Google Cloud Platform, PowerBI
– Spearheaded data pipeline initiatives for the PHVote 2022
election coverage website, encompassing data ingestion,
processing, and analytics using Django, Redis, Python, and
SQL.
– Led organisation’s political dynasty research through
database deduplication and similarity scoring
– Innovated audience engagement with the Moodmeter API and
enhanced article recommendation systems via ontology graph
databases.
– Developed organisation’s ontology graph database to help
improve the site’s article recommendation system and
tagging system using OwlReady and graphDB