Virtual
In-person
How to accidentally lose $2m to API breaking changes (or how to prevent it)
In 2019, Anthony Alaribe led a legacy service rewrite at Delivery Hero which turned out poorly, due to a single missing field that no one knew about. That mistake caused an almost $2 million loss for the company. This talk will explore that journey, including Anthony’s retrospective, and how he devised an algorithm to automatically catch future breaking changes or missing fields across arbitrary APIs to ensure that kind of issue never happens again.
This webinar will be a dive into relevant cryptography primitives, which make it possible to perform such analysis at scale. It will discuss:
- The challenge of legacy service migrations
- Flattening real-time API streams into a form that can be analyzed
- Merkle trees and how they can be applied to catching breaking changes