Berlin Buzzwords 2026: LIVE Recording with Doug Turnbull and Trey Grainger
About their new course for AI-powered search and their views on the search industry in general
It is a particular pleasure to be able to reconnect with search thinkers and tinkerers at a conference like Berlin Buzzwords — which in many ways continues to be the unique, vendor-agnostic, space to meet and exchange ideas that eventually push our industry forward.
This year at Berlin Buzzwords I had the pleasure to record with a few such individuals. This first episode I’m publishing is with Trey Grainger and Doug Turnbull, whose names don’t need special introduction. They are launching a new course targeted at search professionals that are figuring out how to blend “old” search techniques with the “new” agentic AI.
You will find all the relevant links in the show notes of the episode. I will leave you to these two quotes, that among others, stood out for me, as I focus on building a search engine technology these days at Aiven.
The question I asked was: what message would you send to the builders of the search engine tech, that in many ways are working hard on building the infrastructure, but also new algorithms and approaches for improving search?
The managed search engine companies give you the lego pieces for you to go build your solution. And those companies should continue to focus on lego pieces and not on the wholy wrapped solutions, because the field changes too quickly. We haven’t even talked about things like agentic memory, which may be should be using these systems. And also, obviously, it’s kinda taken for granted, but realy robust and highly scalable, “it’s gonna work”, “it’s not gonna have weird edge cases”, is obviously true.— Doug Turnbull
I think it is really important for the vendors to keep in mind this notion of the building blocks, and the ability to plug in these different components, but I also think it is important for us to realize that the definition of what a search system is is not fixed. If I’m building an AI system, and there is an agent — it’s running searches, getting resutls, seeштп if the results are good, rewriting the query and doing it again — where is the search engine? Is it only one of the indexes, all of the separate indexes, the whole system including the LLM? The notion of AI versus search — are kinda merging together. So.. as we are building these systems, we don’t lock ourselves into thinking: search is only this piece, and AI is this piece. Beause when you do that, you don’t build the right types of solutions. — Trey Grainger
Here is the episode on YouTube:
