A seasoned Machine Learning Engineer with a strong academic background, holding a Master's in Machine Learning and Computer Vision and a Bachelor's in Computer Science.
With extensive experience leading the ML division, excelled in developing a reinforcement learning model that revolutionized financial planning efficiency and managed complex data collection and cloud infrastructure.
Previous experience involved training CNN models and applying dimensionality reduction techniques to image datasets.
Proficient in Python, Java, and various ML libraries and tools, combining technical expertise with leadership skills, demonstrated through innovative projects.
Skills AI Consultant, AI Research Scientist, Computer Vision Engineer, Machine Learning Engineer Education March 2021 - Present Master of Machine Learning and Computer Vision at The Australian National University March 2016 - Nov 2018 Bachelor of Computer Science at Swinburne University of Technology Experience July 2019 - May 2024 Machine Learning Engineer at Milk Chocolate – Lead the ML division of our business for the last 5 years– Trained and mentored junior ML engineers– End-to-end development of a reinforcement learning model that generates financial plans for our clients– Developed a custom RL environment that captures all of the logic related to property portfolios– Developed a custom RL model to handle invariable actions– Implemented successful reward, observation shaping– Deployed model to web application, allowing for a manual mode, an automated mode, and a hybrid mode– Final result outperforms manual portfolio produced by our economist and reduced time from 2 days to 1 minute– Implemented scrapers/parsers to collect data for our business, often which exploit complex Javascript-based functionality– Built and oversaw our databases and cloud infrastructure Jan 2019 - May 2019 Machine Learning Fellow at Fellowship AI – Trained headless CNN models on Celeb A dataset of images– Used dimensionality reduction methods PCA, UMAP, ICA, t-SNE to project image clusters in a two-dimensional plane– Developed a metric to identify image type given a set of images #J-18808-Ljbffr