Tradeshift Replaces Legacy BI with Amazon Quick, Achieving 30x Faster Queries and 40% TCO Reduction
Decision Brief
Tradeshift previously used a custom BI tool constrained by a 10,000-row query limit, 25 MB report cap, and 6-month data retention, consuming 50% of engineering resources. After migrating to Amazon Quick, they deployed 16 embedded dashboards, natural language queries, automated workflows (Quick Flows), and deep research (Quick Research) across a three-tier architecture, reducing query times from 45–90 seconds to under 3 seconds. For analysts and operations teams: the internal support team saves 8.5 hours per week on manual CSV reports, external buyers save 6–8 hours per user per week, manual data operations drop 80%, and operational headcount fell from 0.5 FTE to near zero. Infrastructure costs decreased 35%, licensing costs consolidated to a 30% reduction, and advanced analytics contributed 2% ARR growth. Tradeshift also built an AP Auditor chat agent that queries invoices and order status via natural language without coding. The agent connects 11 knowledge bases, 9 dashboards, 14 query topics, and 68 automation tools, explaining metric definitions and data sources, greatly reducing reliance on the BI team.
Sources
- AWS:Machine Learning Blog
Applied ML, infra, and deployment guidance useful for AI builders on AWS.
- AWS:Machine Learning Blog
留言
登入后即可留言,和其他 builder 交换实测心得。
还没有留言,抢头香。