How AI Streamlined New Client Categorisation and Saved a Law Firm Time and Money

Revolutionising Criminal Law with AI

This case study outlines how I helped a law firm use Artificial Intelligence (AI) to streamline new client enquiries, resulting in significant savings of time and money.

Stuart Miller Solicitors, a successful criminal defence law firm based in London, receives hundreds of new client enquiries daily from people being investigated or charged with a criminal offence.

Each enquiry typically includes a complete account of events leading up to the person’s arrest – an example being:

“I’m emailing on behalf of my bruv who’s currently in court for something he ain’t done. Last month he had one too many and this stranger dissed me, so obvs he got a bit mad. We tried to chill it but then a fight kicked off between him and the other geezer. My bruv got nicked and charged, so I wanna know what his options are cos we gotta be back in court in a coupla weeks.”

Due to the complexity of these enquiries, a team of qualified lawyers and paralegals was employed to analyse them.

The team attempted to determine the following:

  • Has this enquiry been sent from the accused or the victim of a crime?
  • For which criminal offence is this person most likely to be charged?
  • Could this case lead to a Crown Court trial (serious criminal offences only)?

The result of this analysis would then determine the speed of the law firm’s response to the enquiry, and which lawyer should respond. A speedy response to high-value cases could generate significant fees for the law firm.

However, this process took up a lot of time and resources, meaning that highly qualified, well-paid individuals had to dedicate their time and energy to processing email enquiries instead of working on client cases.

That’s why I developed an app for Stuart Miller built on OpenAI’s GPT-3 technology. This app receives the content of the client’s email and makes the same determination as the human team of lawyers. It then immediately alerts the relevant lawyer to any enquiries flagged as high-priority, high-value cases. The entire process is automated requiring no human interaction. 

For two months, we tested the app’s accuracy against the evaluations performed by the human team of lawyers and paralegals. At the end of the test period, we determined that the app was processing enquiries significantly more accurately than the human team. As a result, the human team has now been disbanded and released to work on client cases.

Implementing this AI system has resulted in a significant saving in time and money for the law firm, increased its ability to convert high-value clients, and allowed the lawyers to engage in more satisfying and productive work. The system was devised, planned, and launched in just three days. The cost of running and maintaining the app is less than $0.02 per client enquiry.

Please get in touch if you’re interested in exploring how AI could be utilised within your organisation.