Systems | Development | Analytics | API | Testing

Generative AI has completely changed the landscape of enterprise automation, knowledge work and operational efficiency. In 2026, the question is no longer whether these models can perform complex tasks, but whether they can do so reliably enough for mission-critical systems. Despite the availability of sophisticated models and expansive context windows, technology leaders continue to face frustration. Organizations struggle to produce consistent and repeatable results.

Enterprises are under pressure to deliver software faster without sacrificing trust. AI generated code, continuous delivery, and increasingly agentic systems are accelerating change faster than traditional quality practices can validate it. For enterprises running multi-layered tech stacks, weekslong regression cycles and performance issues that are discovered by customers in production are symptoms of a behind-the-scenes quality model that was built for a slower era.

We audited our skills library a few months ago and found twelve dashboards hiding in it. Not dashboards. Skills that built dashboards. Someone needed a view of some data, asked Claude to put it together, got a long HTML page out of it, and then wrapped the whole thing in a skill so others could run it again. Twelve times over, by different people, for different questions.

Last week we introduced AI Transport v0.2.0 and made one idea the centre of the design: the session is the channel. Every input, output, and lifecycle event for an AI conversation is just a message published to an Ably channel, which is what makes a session durable, multi-party, and resumable. In v0.3.0, we added first-class support for presence and LiveObjects to AI sessions, allowing you and your agent to see who's online and update shared state in real time.

Hyper-personalization in 2026 is the ability to act on a user's current intent within the current session, using signals from across the journey. Batch customer data platforms (CDPs) can't do this. They can't capture intent as it forms, can't hold session state, and can't activate inside the intent window.

The problem. Predictive ML pipelines that maintain separate batch and streaming code paths for the same features carry training-serving skew, the gap between the features a model was trained on and the features it sees at inference time. Skew silently degrades model accuracy and doubles infrastructure cost. The recommendation. Adopt a unified streaming (kappa) architecture.

Your dashboards have a new reader, and it doesn’t have judgment.