Ryan Siegel
Data engineer and analytics engineer. Boston area. 13+ years in data.
About
I build and test dbt models on Snowflake for MFS's investment data platform, a 3+ year project that has users today and is still rolling out features for broader use. Portfolio managers and other internal teams use the data for investment decisions.
Earlier, I was an analytics engineer at WorkHuman, where I wrote the dbt guidelines and automated analytics refreshes in AWS Glue. Before that I was a senior data analyst at WorkHuman and Robin Powered, and I spent six years at Criteo on advertising analytics.
I hold an M.S. in Data Science from Northeastern University.
Projects
Projects in progress
Two data engineering projects are going up on GitHub soon.
Experience
MFS, Boston
- Build and test dbt models in Snowflake for the investment data platform, including intermediate models behind the consumption layer.
- Own failure monitoring and remediation for 1,900+ dbt Cloud tests on Snowflake models.
- Led a roughly six-month health checks project, directing two contractors on dbt models that compare raw data with platform output across 12 asset types.
- Redesigned the dbt release process, moving releases from about every three weeks to weekly, so support now runs them without engineers.
- Built the Data Refresh Validator, which replaced about 5 hours of manual command sequences with about 10 minutes of input.
WorkHuman, Framingham
- Wrote the dbt guidelines and trained all seven analysts through workshops and ongoing support.
- Automated a manual weekly refresh of about 30 analytics tables in AWS Glue, cutting manual work from 25 hours to 1 hour a week.
- Introduced controlled A/B testing in place of sequential before-and-after comparisons; a 30-day chi-square test showed a 10% discount lifted sales about 7.5% on average.
Robin Powered, Boston
- Cut month-end board reporting from 20 days after close to 5 by automating processing and transformations in Python.
- Defined revenue metrics (churn, contract upgrades, new versus existing customers) and modeled Salesforce data into a monthly Redshift table; the work showed 12% of the business was churning, which was presented to the CFO for the board.
Criteo, Boston
- Ran dozens of client A/B tests, mostly incrementality tests, that showed Criteo's value and guided client budgets.
- Built and presented Black Friday through Christmas analyses to directors, which informed budget increases.
- Wrote the SQL behind a campaign dashboard refreshed every five minutes that let a commercial team of about 16 adjust CPCs through Black Friday; reused for LATAM and Brazil.
- Temporarily transferred to help ramp up the Chinese market, training local teams and applying U.S. analytics processes.
- Built "Beyond the ROAS," a Python analysis adopted for 80% of North and South America client QBRs.
- Automated client reporting with Python scripts that populated PowerPoint and Excel directly, eliminating manual copy errors.
- Parsed retargeting campaign data to give client managers actionable intelligence.
Skills
Core: SQL, Python, dbt, Snowflake, data quality and testing
Also: AWS Glue and Athena, Azure, Fivetran, Tableau, Looker, Git and CI/CD