Meet Faraaz Sheriff

Before I started Infomathics AI, I spent 12 years building data and AI systems end to end — from ingestion and pipeline architecture through machine learning models, production deployment, and the reporting that stakeholders actually use. I worked across some of Canada's largest organizations and global enterprises: building predictive models for Johnson & Johnson, designing churn and pricing analytics for TELUS, architecting a demand forecast used by Southwest Airlines that generated an estimated $1.9M in annual savings, and spending three years inside the Ontario government integrating sensitive, siloed health and social services data that now underpins policy decisions for millions of people.

What that experience taught me — more than any particular technology — is how hard it is for organizations to actually operationalize intelligence. You can build a brilliant model and still have it ignored by the people it was built for. You can implement a state-of-the-art system and still have teams refuse to change how they work. The gap isn't technical. It's human: the skills, the workflows, the leadership buy-in, and the practical understanding of what AI can realistically do in a given context. That is what Infomathics AI was built to address.

I work with executives who need to make smart AI investment decisions, with operations teams that want to automate repetitive work, and with managers who need their people to actually use the tools their companies are paying for. I train teams on the real tools — Claude, ChatGPT, Microsoft Copilot, GitHub Copilot — in the context of their actual workflows. And when training surfaces an automation opportunity, I can build it: I have 12 years of production engineering behind me.

The Engineered Supply project is a good example. It started as an AI training session for their team. What surfaced during that session was a documentation process eating 1–3 hours per project across hundreds of projects a year. We built a system to automate it completely — zero errors, every logbook generated to spec, delivered as a one-click workflow. That is what I mean by measurable business impact.