EdTech Data Architecture
Data foundations and marketing intelligence for Ampli, within Cogna Education
Business need
The company had just launched Ampli, a fully mobile learning platform offering online graduate programs, professional upskilling, preparatory courses, and short specialization tracks. Early adoption was strong, but the business had no data architecture to support growth. There were no unified indicators, no consistent data definitions, and no automated flow from the operational platform to analytical environments. Leadership could not see daily enrollment trends, revenue trajectories, churn risk, or the performance of each course or acquisition channel.
Student journeys were also invisible. It was impossible to track how users progressed through onboarding, where they dropped out, what content they engaged with, or how demographic profiles influenced learning behavior. Personalization was not feasible, since the data was fragmented across multiple systems that did not communicate. The company’s shareholders believed in the product and saw qualitative signals of success, but the absence of reliable dashboards and clean data pipelines made it impossible to demonstrate the financial impact or manage Ampli with confidence.
My Role and Contributions
I designed the end to end data architecture for Ampli, starting from how raw events were captured in the platform and ending in a clean model connected to our management system (SAS). I mapped every operational table, defined the business rules for enrollment, cancellation, revenue recognition, and student status, and then structured a unified database that could support both financial reporting and marketing intelligence. In total I created more than thirty rules and indicators, including cohorts, active base, lifetime value, churn, conversion along the funnel, and engagement scores at student and course level. The pipelines were built with SQL, Power Query, VBA, and Python, with automated refresh and data quality checks to keep the numbers consistent.
On top of this foundation, I worked with Finance, Product, and Marketing to translate their questions into data products. I delivered executive dashboards that fed the company’s quarterly earnings releases, including revenue, margin, and intake performance for Ampli. For the growth team, I created segmentations that combined demographics, channel of acquisition, and behavior inside the app, which supported targeted campaigns and experiments. I also documented the model, trained analysts on how to use it, and set governance routines so new courses and features would enter the pipeline correctly. The architecture I built continued to be used for more than two years after I left the company, and later served as the blueprint when the entire data environment was migrated to AWS, with my original data model as its foundation.
