Reflecting On the Journey
Recently, I was able to stamp 'project complete' on a remarkable collaboration with Google, leading a project team to analyze and advise on Shopify’s big data and marketing technology (MarTech) approach. Core to this deliverable was utilizing advanced machine learning and cross-platform techniques to identify relationships across big data.
It was an excellent opportunity to flex some AI/analytics muscle and work with such a fantastic team. It’s not every day that one gets the opportunity to pair up with Google and present insights globally to senior practitioners in marketing, analytics, platform / digital transformation and creative.
Reflecting back, there are a couple key principles that really rang true. If you’re passionate, you don’t count the hours, but rather, the marks of progress. This project was less about completing the task and more about feeding our curiosity, as we uncovered trends, which shaped into findings, and then ultimately, manifested into insights that we recommended to the client. However, this was by no means a linear path. It required significant exploration, feature engineering and experimentation, where our route to success was highly dependent upon openly embracing the scientific method. Each step forward, no matter how small, celebrated as progress.
“If you’re passionate, you don’t count the hours, but rather, the marks of progress.”
Looking at the second principle, we discovered that the value of working within a fully integrated data pipeline is worth its weight in gold when taking on big data. We were faced with the task of collecting, extracting, consolidating, engineering, analyzing, and finally, reporting on multiple datasets from various sources that were gigabytes in size, contained millions of entries (instances) and had over a hundred thousand attributes. This was no small task. By design, we frontend loaded our efforts investing 60 percent of our time towards data engineering; moving data frequently between BigQuery and Google Colab. For those curious, our full pipeline leveraged Google Cloud Platform, Cloud Storage, Cloud Video Intelligence API, Vision API, Speech API, BigQuery, Google Data Studio, Google Colab, and even Google Slides, to tell the story.
Thank you, Smith School of Business, for the trust, opportunity, and belief in the process. This was an opportunity established by Queen’s University and selectively offered to only a handful of graduate students as part of the Master of Management in Artificial Intelligence (MMAI) program.
