Executive Bio
I lead applied AI and data science programs by connecting strategy, experimentation, and production delivery. The focus is always the same: make the work trusted enough to change decisions at executive level.
My work sits at the intersection of AI strategy, operational rigor, and commercial accountability. Over the last 12+ years, I have built systems that help large organizations move from analytical possibility to repeatable business action.
I have led data science across forecasting, personalization, experimentation, and ML-enabled decisioning in retail and fintech settings. The through-line is consistent: models only matter when the surrounding organization can trust, operationalize, and measure them.
That is why I focus as much on decision layers, operating cadence, and cross-functional alignment as I do on the technical stack. Durable value comes from the system, not from isolated technical wins.
The job is not to produce clever model outputs. It is to make those outputs usable, trusted, and operational inside real teams.
Mature machine learning programs win when delivery becomes dependable enough for the business to plan around it.
I build teams that can prove impact with experiments, operating metrics, and clear investment logic.
Most organizations invest heavily in model development, then underinvest in the layer that translates scores into action. My framework for closing that gap has three parts.
Produce trustworthy model outputs with clear business intent, not generic predictions disconnected from use cases.
Define the rules, constraints, and prioritization logic that convert signal into a business action teams can actually execute.
Measure outcomes, inspect reliability, and feed what is learned back into roadmap and organizational decisions.
Amir Tavasoli is a Director of Data Science at Sobeys, one of Canada's largest food retailers, where he leads advanced analytics initiatives focused on transforming marketing through machine learning and agentic AI.
He began his career in machine learning after earning a Master's degree in Computer Science from McMaster University. His research focused on developing a natural language processing engine to triage communication between patients and physicians, and was presented at leading conferences including AMCIS and IBM CASCON.
Amir has over a decade of experience applying data science across retail and fintech. He spent six years at Canadian Tire, driving improvements in forecasting and inventory optimization, followed by a role at The Home Depot's eCommerce team, where he advanced search optimization and recommender systems. He later served as Machine Learning Lead at Paytm, leading large-scale personalization initiatives across the platform.
At Sobeys, Amir has played a key role in building and scaling personalization capabilities in marketing. Today, he focuses on delivering measurable business impact by enabling AI-driven decision-making and accelerating adoption of AI-powered solutions in Marketing.