👋 Hey, I'm Paul. I am a machine learning practitioner based in Toronto. Over the past 5+ years, I have been designing and building intelligent applications and web experiences. I am a builder at heart, and my passion for making an impact in society drives my work.
I have strong technical background in Python, TensorFlow, Dataflow and Spark execution engines, and have successfully applied deep learning techniques to a variety of domains including computer vision, natural language processing with transformers, data analysis and machine learning for competitive data science.
Presently, I focus on supporting businesses in executing large scale analytic programs to achieve their full potential. As a Cloud Data and ML Engineer, I implement scalable, secure, and cost-effective cloud-based solutions that meet the needs of the business.
Examples of my recent work include
Use case in global anti-money laundering: Reduced the risk of money laundering and fraud offences through big data analytic solutions that ensure rules for revealing fraud and money laundering evolve in step with offenders practices.
Use case in digital advertising: Growing acquisition volumes, value and efficiency leveraging standardized processes and technology capabilities for building, deploying, and operationalizing ML systems to maximize return on advertising spend.
Let's build together! Connect with me on LinkedIn.
Ideas, analysis and strategy.
Exploring the art of AI/ML, one concept at a time.
Advanced hyperparameter optimization techniques including Bayesian, Multi-fidelity, and SMBO algorithms.
A graph processing platform to create a ranking dashboard of influential Twitter profiles using Neo4J and Spark.
A neural network model for predicting CLTV scores using Tensorflow and Lifetimes Python Library.
An advanced Twitter scraping & OSINT tool written in Python.
Weekend hacks for fun and learning.
A better Hacker News.
A curated list of my favorite tools and software.
Tools and resources for staying safe on the internet.
Internet things, saved for later.