Over the past few weeks, my world has shifted — from Jira boards and bots to Jupyter notebooks and joint distributions.
From driving Robotic Process Automation at TD to delving deep into machine learning models, Unix scripts, and data ethics at the University of Toronto’s Data Science Institute — it’s been a ride.
Why share this now? Because in making a mid-career pivot into AI, I’ve found myself sitting with a tough but honest question:
How do I make sure what I’m learning actually drives business value — not just model performance?
Coming from a product and automation background, I’m used to asking “Why are we solving this?” But in machine learning, the conversation often starts with “How accurate is it?”
That disconnect caught me off guard — and also lit a spark. It made me realize: this is the space I care about most. Bridging logic and impact, models and meaning.