Challenge

Lime used a third-party ETL pipeline (extract, transform, load) to export Stripe data to Snowflake and noticed that the data provided by the ETL pipeline wasnt always matching up with Stripes data. A data engineer would have to dive into the cause of each discrepancy, eating up to two hours of their time for each inconsistent transaction. Since Lime was using a third-party ETL, they could not fix the root cause of these discrepancies and realised that they needed a more reliable solution to stream Stripe data to Snowflake.

Solution

Lime chose to adopt Stripe Data Pipeline, which automatically sends the complete Stripe data set to its Snowflake Data Cloud on an ongoing basis. By adopting Stripe Data Pipeline, the company saved weeks of engineering time as it did not have to build and maintain an API for every single Stripe data set. Out of the box, Lime was able to access its current Stripe data directly in Snowflake.

Results

Since implementing Stripe Data Pipeline, Lime has been able to automate and speed up its financial reporting. Engineers no longer need to look into data discrepancies as all transaction data matches up with their reports, allowing the team to focus on product initiatives and speed up product velocity.

The data quality and data freshness of Stripe Data Pipeline are a welcome departure from third-party ETL pipeline vendors, says Nabil Barakati, Software Engineer at Lime. Data imported through Stripe Data Pipeline always matches our reports exactly. Where it would take three days to get transaction data via an ETL pipeline, it only takes one day with Stripe Data Pipeline.

Lime is also benefiting from richer business reports with access to the complete Stripe data set. Limes finance team can now track refunds in nearreal time, which allows them to confidently report on the companys latest cash status. Limes fraud team is creating improved fraud models, as they can now merge fraud signals from Stripes risk scores per transaction with their proprietary fraud data.