99.9% uptime
Kept Canada’s largest auto-insurance policy-history database fast and available, and built the data marts on top.
Context
A 2 TB SQL Server database of automobile-insurance policy history, with finance, analytics and delivery teams depending on reports built from it each month.
What I did
I managed the database and built production ETL/ELT pipelines for 22+ micro data marts, each under six tables and on a monthly SLA.
Outcome
99.9% uptime. Median queries under 200 ms. Report runtime cut from 30 minutes to 5.
Evidence
Client work, no public page.
Stack
SQL Server, T-SQL, PostgreSQL, Python, ETL/ELT