Snap One (acquired by ADI Global) · 2021–2025
The Metrics Behind a 50% EBITA Increase
Built the Databricks pipelines and the governance around them that took IoT and SaaS telemetry all the way to board-level decisions, contributing to a 50% EBITA increase and a multi-million-dollar ARR win.
- Databricks
- SQL
- Python
- Power BI
- ETL / ELT
- CI/CD
- Heap.io
- EBITA increase (~$50M) the pipelines contributed to
- 50%
- Databricks pipelines integrating IoT and SaaS sources
- 15+
- lift in executive adoption of BI reporting
- 60%
- faster query execution across 100+ standardized reports
- 40%
EBITA increase (~$50M) the pipelines contributed to
Databricks pipelines integrating IoT and SaaS sources
lift in executive adoption of BI reporting
faster query execution across 100+ standardized reports
The problem
Product, sales, and finance were all making decisions from the connected-device estate, and all reaching different numbers. The data existed; what did not exist was any reason to believe it. Every executive review spent its first half arguing about whose figure was right, which is the most expensive possible way to run a meeting and the surest way to have good analysis ignored.
The approach
I developed and automated 15+ Databricks pipelines integrating the IoT and SaaS sources, then standardized and optimized 100+ SQL queries and BI reports across Power BI and Databricks. The durable part was governance rather than pipelines: anomaly detection, CI/CD-enabled data quality checks, version control, and a gold-metric certification step, so a number reaching a dashboard had passed something. Alongside the engineering I served as Scrum Master for two geographically dispersed US and EU data science teams, and as Data Product Owner ran vendor selection and integration for the Heap.io product analytics program.
The outcome
The pipelines contributed to a 50% EBITA increase (~$50M) and a multi-million-dollar ARR customer win. Query execution got 40% faster and executive adoption of BI reports rose 60%. But the real change was that dashboards on outages, sales trends, and customer segmentation started informing decisions the product managers made and the ones the board made, because certification made them arguable on the merits instead of on provenance.