Top News

Cloudflare has introduced Cloudflare Radar Researcher, a new AI powered research tool made to help people sift through the company’s huge stash of Internet observations. It was announced on August 7, and basically it lets users ask about Internet patterns in normal language, then get answers that come with supporting data plus interactive visual views.

Radar Researcher is built on top of Cloudflare Radar, which is the company’s public platform for watching Internet traffic, outages, security themes, routing patterns, and other changes across the global network. Instead of forcing someone to go and locate the correct Radar page, then decode separate metrics, or create complicated API queries, this new conversational layer tends to do a lot of the heavy lifting automatically.

Cloudflare Radar data becomes conversational.

Cloudflare launched Radar back in 2020 to provide visibility into global Internet trends, drawing from the information collected across its network. The platform’s dataset is used by many different kinds of groups, including journalists, researchers, human-rights organizations, network operators, and academics.

Still, pulling out one particular insight can end up feeling a bit hard; users have to figure out where the relevant data actually lives and how the different metrics are meant to behave or, well, what they really measure. Radar Researcher exists to take away that extra friction. People can just type a question, like “how is Internet quality in this country,” and move on.

The AI then decides what data is needed, hits the right Radar APIs, sorts through the findings, and returns something that reads like plain language. Also, the output is not limited to plain text only. Radar Researcher can produce interactive charts in much the same spirit as the ones you might already see in Cloudflare Radar, so the underlying evidence is easier to trace and check.  

Users can ask follow-up questions.  

One of the more useful parts of Radar Researcher is its conversational approach. After someone gets an initial answer, they can ask follow-up questions without having to rebuild the whole analysis from the beginning. For instance, after looking at Internet quality in one country, a researcher might ask for a comparison with a neighboring country.

Radar Researcher then retrieves new data for those additional locations and keeps going with the analysis. On top of that, the system can suggest likely next investigations on its own, sort of nudging users toward relationships or trends they might not have thought to look for in the first place. This is why the product feels more like an AI research assistant, rather than a conventional analytics search box.

Final Thoughts

Cloudflare Radar Researcher feels like a pretty practical use of generative AI for Internet analytics. Instead of just making yet another standalone chatbot, Cloudflare basically hooked an AI agent straight into the structured data and visualization stuff that Radar already has.

So you end up with something that can take questions about Internet traffic, connectivity, outages, and network trends and then give data-backed answers plus interactive charts. For researchers, and honestly most other people who regularly rely on Internet data, the main shift is pretty simple: you don’t have to learn how to query the data first; they can gradually start with just asking the question.