Skip to content
Security & Trust Talking point

Fingerprint says its AI connector lets fraud teams question bot data without SQL

A Fingerprint speaker described an AI connector that lets teams probe bot attacks without writing database queries.

W
WebPulse Newsroom
AI-assisted · 2 min read
Share on X LinkedIn
Fingerprint says its AI connector lets fraud teams question bot data without SQL
In brief
  • Keshia Rose of Fingerprint said its AI connector lets people investigate bot attacks and device clusters in plain language, without writing SQL.
  • In our reading, the hard part moves to judging whether the answer is right, which Rose did not discuss in the excerpts we reviewed.

Keshia Rose of Fingerprint said the company's AI connector lets fraud teams question their bot-attack data in plain language instead of writing SQL. She made the point on Software Engineering Daily, in an episode on fraud prevention and AI agents. Rose works in developer relations at Fingerprint, which sells tools that identify visitors to a website.

What was said

Rose said Fingerprint built an MCP server. MCP is a standard way to connect an AI tool to a product's data and functions. She said its main use is analysing your own Fingerprint data.

Her examples were investigative. The tool can help you work out what happened in a bot attack and where it came from. It can also show whether some clusters of devices are doing more fraud than others. She said this works in natural language, so nobody has to open SQL and run their own analysis. SQL is the standard language for querying databases.

She said Fingerprint plans to expand this. It has also launched a command-line tool and a page for installing the product through AI tools. Her reasoning was practical: "It doesn't matter if you're a vibe coder or if you're a senior engineer, people are using AI tools."

Why it matters

In our reading, writing the query used to be a gate. Someone had to know the data layout and the language before asking whether a cluster looked suspicious. If a plain question can replace that step, more people can start an investigation. That holds for Fingerprint's data, as Rose described it.

The harder step comes after the answer. Someone must decide whether a flagged cluster is real fraud or an ordinary pattern. Someone must also judge whether the question was framed well. Managers who buy such tools should ask who on the team can do that. They should also ask how newer analysts will learn the skill if they never write the queries.

The other side

In the excerpts we reviewed, Rose did not discuss checking the AI's answers or what happens when it is wrong. She works for the company selling the tool. Nothing in the excerpts shows how accurate the answers are, or how any team has changed its training.

She also did not say that analysts' skills are shifting. That is our interpretation. The excerpts leave open whether natural-language analysis helps experts work faster, or lets less experienced people reach conclusions they cannot yet test.

Written by the WebPulse Newsroom with AI assistance, and checked by our editorial review: every quotation was verified against the recording's transcript. How we use AI.

The conversation this talking point comes from