Case study
Pharmaceutical Manufacturing
Snowflake
840M
Records Eliminated
The Challenge
Snowflake costs climbed relentlessly, driven by 840M unnecessary records inside a 1.2B-row environment serving over 60,000 users, with no cost ownership or governance.
Seventeen warehouses ran on inefficient default configurations, with long idle times, expensive queries, and ETL jobs reprocessing entire datasets on repeat.
Fragmented ownership across BI and IT, plus inefficient Sigma and DBT workloads, produced redundant processing and near-zero spend visibility.
Our Solution
Optimised Snowflake by tuning expensive queries, configuring warehouse suspend/resume policies, converting staging tables to transient tables, implementing clustering keys, and enabling proactive cost monitoring.
Ran AI-driven discovery to identify unused tables, archive obsolete data, detect redundant DBT and Sigma workloads, and surface warehouse consolidation opportunities.
Conducted an end-to-end architecture assessment mapping the full data pipeline, eliminating bottlenecks and producing a prioritised optimisation roadmap.
Business Impact
Cut Contact ETL query runtime from approximately 15 minutes to under 30 seconds while eliminating 840 million unnecessary records.
Improved warehouse efficiency by reducing idle suspension time from 60 minutes to 60 seconds, with real-time dashboards giving continuous cost visibility.
Established a roadmap to consolidate 17 warehouses into 6–8, strengthening governance, cost accountability, and Snowflake operating expense.
Workato ETL: How Scheduled Product-Data Sync Cut a Week-Long Process Across ~70 Brands
Explore Story
Salesforce Transformation: How AI-Powered Metadata Analysis Untangled Fragmented CRM & CPQ Environments
Explore Story
Salesforce Service Cloud: How a 360° Patient View Replaced Siloed Support Systems and Legacy Tech Debt
Explore Story
