Introducing Pia, Your Own Pirical Insights Analyst

Aidan Scanlon • 13 October 2026

3 minute read

Pia is a Pirical Insights Analyst on your team. Ask Pia to match lawyers to a job description, map a market in seconds, or turn a question into a briefing you can take into the room.

When you search for a role in Claude, ChatGPT, or Copilot today, at first glance the answer looks good. This is because these tools search the way you would search the web. They look for keywords, pick out profiles and return what seems to match.

But because there is no database behind them, the search is superficial and misses around 80% of the market, and 40% of what comes back doesn’t match the job spec.

Pia is different, it works from a database that we’ve vetted for over a decade. To put this to the test, we've evaluated Pia's completeness and accuracy against the most common LLMs on the market by comparing the answers from a sample of real user searches.

Here are the results!

Pia returns 4x more profiles from attorney searches than standard LLMs

For every lawyer a standard LLM finds, Pia returns 4. That means 8 out of every 10 attorneys are missing from tools like Claude, ChatGPT, Gemini, and Copilot.

Pia's results are 1.5x more accurate than standard LLMs

Data completeness and accuracy across each model varies, but on average a standard LLM only returns 20% of relevant profiles, with an accuracy rate of just 63%.

"This would elevate conversations with Partners. It allows us in Talent Acquisition to have a much more commercial discussion about the market."

Talent Acquisition Specialist

Clifford Chance

"It is brilliant! The market insights it provides are so helpful. This has already saved me a lot of time."

Recruitment Manager

Legal 50 law firm

What makes Pia different?

AI is only as good as the data it's built on. We've been collecting and vetting the data you find in Pirical Legal Professionals for over 10 years. Pia answers your questions by referencing this data, no open web searches or best guesses.

Pia can draw upon:

0 lawyer profiles

0 in-house counsel

0 client matters

Where standard LLMs fall short

It you ask Claude, ChatGPT, Gemini, or Copilot to find you 'Chicago-based Real Estate lawyers with 10+ years of experience', you will receive a list of candidates that appear relevant. But how complete and accurate will that list be?