Database and infrastructure

One AI search across ScyllaDB's many product docs sites

ScyllaDB documents each product on its own site. AI search unifies them into one box: developers get fast, ranked results across every site, and can fall back to an AI answer when a quick search does not surface it.

Biel has transformed the way users navigate the ScyllaDB documentation, giving them quick results when they know what they’re looking for and helpful guidance when they don’t. It makes discovering the right information faster and more effortless, reducing the friction of navigating a large and complex documentation set.

The challenge

ScyllaDB is a high-performance, low-latency NoSQL database, and its documentation is not one site but many. The database manual, Cloud, Monitoring, Manager, Operator, and the drivers each live on their own site.

Developers working at scale need to find the right page fast, and it might live on any of those sites.

How ScyllaDB deployed Biel

ScyllaDB made AI search the primary entry point, with one index spanning all its product sites. A developer types a few keywords and gets ranked results instantly across every site. For experienced users that speed matters: a quick keyword is often faster than waiting for an AI chatbot to answer.

When a quick search does not surface what they need, the same box falls back to AI: the top result offers to answer the question directly from the docs, with sources. A full Ask AI chat is available too, for longer questions.

What the search log reveals

The search log is a live map of what a database audience needs to find fast, ranked by how often developers look for it.

For the docs team, that ranking is a direct signal of which topics to keep sharp and where to invest next.