Systems | Development | Analytics | API | Testing

I spoke about Ably's approach to resilience at the AWS Cloud Resilience Roadshow in Dublin on 22 September 2026, and the questions afterwards covered CDNs, regional failover, outages, message ordering, and capacity. They got me reflecting on which parts of our approach are common across the industry, which are less so, and how choices made early, well before my time at Ably, have put us in a strong position today.

Most mainframe storage refreshes begin with a notice. A support statement gets revised, a lifecycle date is published, a lease comes due, or a capacity report shows the estate crossing a threshold two years earlier than anyone planned for. A purchase that had been sitting comfortably in a future budget cycle now has to close in this one. Under deadline pressure, the instinct is to buy the current generation of the platform you already run.

Every AI conversation eventually turns into a model conversation: which one is fastest, which one hallucinates least, which one your competitors are already using. Three recent episodes of The Data and AI Chief, hosted by ThoughtSpot's Chief Data and AI Strategy Officer Cindi Howson, make a different case entirely about what it takes to scale AI in the enterprise.

Displaying a list of items from the database is probably the most common functionality built for web applications. This is where pagination with LIMIT clause comes into play, which is most likely what you learned to use. This pattern is called offset pagination. In this post I am going to contrast it with cursor pagination and the performance considerations that come with the choice of pagination algorithm. Examples are with Plain PHP and Doctrine ORM 3.7, which ships a new CursorPaginator class.

Traditional CI/CD pipelines relied on determinism, assuming code produced an exact, repeatable output. Generative models and agentic workflows break this model. A single prompt edit or model checkpoint update can pass standard unit tests while causing silent regressions like truncated API payloads or missing parameters five steps downstream. To ship probabilistic AI reliably, engineering teams must redefine what counts as a passing test.

For many teams, Bitrise Insights means build duration and failure rate. Those are good starting points to evaluate your CI performance, but they often fall short when it comes to understanding questions about speed and quality.

Plain-text logs are great if you want a simple explanation of how a function is performing or why a crash happened. But as your application grows, they become difficult to search and analyze. Structured logging makes production data easier to filter, correlate, and use across modern logging tools. This guide will give you the springboard to build structured logs effectively, showing you how to.