How KLog Saved $200,000 by Switching to Bemi Audit Trails
KLog democratizes international cargo transportation with an intuitive digital platform. In addition to moving cargo, KLog offers a comprehensive solution for more efficient and hassle-free logistics management. With over 5,000 customers, including Hugo Boss, Crocs, Kensington, and Wrangler, KLog has become the de facto logtech in Latin America and around the world.
KLog was spending upwards of $200k in engineering resources to track data changes internally
KLog required a highly reliable data audit trail that included context such as the user ID, API endpoint, request payload, and GraphQL mutation name behind a change.
The company initially built a solution internally through a cross-collaborative effort between software and data engineering teams. The system involved first adding an updatedBy field on every Postgres table and creating an application layer middleware that set this field on every data change in serverless functions. From the AWS RDS instances, their team then generated logs using an AWS DMS task and, due to the sheer volume, attempted to store the data in Parquet files for post-processing in S3 buckets. An alternative approach companies consider when building DIY is using Debezium for change data capture.
The system eventually encountered breakdowns when it came to reading the data, ensuring consistent application context, and maintenance.
The DIY system got so complex that developers mentioned they were 'allergic' to that part of the codebase. There could have been lost upmarket deals since it took months to initially build.
— Álvaro Serrano, CTO KLog
KLog switched to Bemi and never looked back
KLog easily connected their Postgres databases to the Bemi platform and used the Prisma ORM integration to automatically handle the enrichment, structuring, and formatting of the lower-level database events from the Write-Ahead Logs. KLog was then able to use Bemi’s intuitive control plane UI to significantly reduce data volumes by tracking only relevant data changes through column and table-level filtering.
Due to the high reliability and accuracy of the platform, Bemi became the 100% source of truth at KLog. Beyond just an audit trail, KLog’s more than 40 customer success and operations team members use the Bemi UI multiple times a day as the definitive, immutable source of truth when troubleshooting shipment updates.
We love working with the Bemi team. Their customer service is incredible — responsive, knowledgeable, and always willing to go the extra mile. They’re a talented team, and we have full confidence in their expertise.
— Álvaro Serrano, CTO KLog
Future plans
Since the Bemi data is also easily consumable with ORM-specific libraries, KLog plans to later build products centered around the data, such as customer shipment feeds. With more internal AI applications injecting data into KLog’s platform, their operations teams are becoming auditors of data rather than data inputters, thanks to Bemi. Bemi plans to extend ORM integrations with functionality to consume data change events in real-time, allowing KLog to reliably also power their notifications, AI RAG system, and microservice communication in the future.
Bemi has been a game-changer for us, highly recommend!! We’re not in the business of tracking data changes and are now able to concentrate fully on our core logistics product.
— Álvaro Serrano, CTO KLog
Try out Bemi
If you want to use Bemi to track Postgres data changes, star Bemi on GitHub and try Bemi Cloud for free.