下载PDF
Kyvos Insights > 实例探究 > E-waste Recycler Transforms Operations by Analyzing 4 Years of Kiosk Data
Kyvos Insights Logo

E-waste Recycler Transforms Operations by Analyzing 4 Years of Kiosk Data

技术
  • 分析与建模 - 大数据分析
  • 分析与建模 - 实时分析
  • 应用基础设施与中间件 - 数据交换与集成
适用行业
  • 回收与废物管理
  • 零售
适用功能
  • 商业运营
  • 销售与市场营销
用例
  • 预测性维护
  • 质量预测分析
  • 供应链可见性(SCV)
服务
  • 数据科学服务
  • 系统集成
挑战
The e-waste recycling company was facing several analytical challenges on their Snowflake / AWS / Tableau platform. They were experiencing severe degradation in query performance while analyzing multiple years of data in Snowflake. There was no semantic layer to define consistent data models and standardize KPI calculations. The company was also dealing with unpredictable querying costs on their Snowflake data warehouse, with bills running very high at times. They found it difficult to model their data and deal with multi-level hierarchy and one-to-many joins between facts.
关于客户
The customer is a leading e-waste recycling company that aims to reduce electronic waste by making it easier for users to sell their unused devices. They have placed automatic kiosks across thousands of retail outlets in the US and Europe, allowing users to drop their devices, negotiate prices, and receive on-the-spot payment within a few minutes. The company is committed to creating a greener planet and improving business efficiency. They wanted to analyze four years of kiosk transaction data to support decisions such as where to place the next kiosk, how to reduce wait times, and how to tailor quotes based on user acceptance rate.
解决方案
The company chose Kyvos to build OLAP on AWS with Snowflake as the data source. Using its Smart OLAP™ technology, Kyvos built a high-performing Smart semantic layer™ directly on their S3 platform. It pulled data from Snowflake and created optimized OLAP models with complex schemas and advanced calculations. Once Kyvos was deployed, query response times were almost instant, enabling their users to deepen analytics across several dimensions. A unified semantic layer helped define all their business logic in one central place, and all BI tools could connect to this layer instead of everyone doing their own calculations. As all the aggregates were calculated in advance, ad hoc queries returned in less than 5 seconds, even on four years of data. Kyvos’ build-once-query-multiple-times approach helped them control budget and reduce querying costs on Snowflake.
运营影响
  • Query response times were almost instant, enabling users to deepen analytics across several dimensions.
  • A unified semantic layer helped define all their business logic in one central place, and all BI tools could connect to this layer instead of everyone doing their own calculations.
  • Ad hoc queries returned in less than 5 seconds, even on four years of data.
数量效益
  • 5-second response time for ad hoc queries
  • Significant reduction in querying costs on Snowflake
  • Instant query response times

相关案例.

联系我们

欢迎与我们交流!

* Required
* Required
* Required
* Invalid email address
提交此表单,即表示您同意 IoT ONE 可以与您联系并分享洞察和营销信息。
不,谢谢,我不想收到来自 IoT ONE 的任何营销电子邮件。
提交

Thank you for your message!
We will contact you soon.