Harvard law student quits to raise $6 million for AI tool that guides police in real time

David Lawrence was midway through his JD at Harvard when a campus incident involving a police officer sparked public controversy over police conduct. Recognizing that many on-the-ground errors stem from officers lacking immediate access to departmental procedures, he decided to build an artificial-intelligence solution. Together with Amit Patankar, a Harvard MBA graduate and former Google engineer, and Michael Grofman, a retired deputy commander of the Boston Police Department, Lawrence founded Blue Voice, a Boston startup that provides real-time policy guidance to officers in the field.

Before Blue Voice, an officer who could not recall a specific protocol—such as the seventh step required at a crime scene—had to flip through a 15,000-page manual, wake a commander in the middle of the night, or search generic tools like Google or ChatGPT. According to Lawrence, consumer-grade models deliver incorrect answers in 30 percentage points of cases because they are not trained on local statutes, municipal ordinances, or internal department procedures. Officers therefore make decisions from memory or educated guesswork, despite having thousands of pages of legislation and policy behind them.

Blue Voice is trained on the exact laws, regulations and protocols of each department—information that general models cannot retrieve from the open web. Rather than generating answers itself, the platform links directly to the relevant regulatory source, leaving the final decision to the officer who combines it with field experience. The system also offers instant access to detailed school maps on a phone during active-shooter events. Lawrence says the platform now answers a query every minute, and officers trust it precisely because it cites sources instead of “making up” answers.

In the past year the customer base grew elevenfold, and today 225 county agencies in 25 states use the tool daily. Lawrence reports a drop in crime and fewer operational disputes. In one incident, a rookie officer identified a man trying to persuade a child into his car but was unsure whether he had legal authority to intervene; the app confirmed the situation met the “child enticement” criteria and granted immediate authority to act. In another case, the system reminded a department commander that an officer cannot return to operational duty after a shooting without a third-party psychiatric evaluation.

Blue Voice competes with Lexipol, which is backed by private-equity investors, and is now adding capabilities to help detectives solve cold cases. As police use of AI—such as Flock Safety’s license-plate tracking systems—faces growing criticism, Lawrence hopes Blue Voice will demonstrate that AI can be deployed for public safety without sacrificing civil liberties. Having previously worked for Connecticut Governor Ned Lamont, Lawrence says he has no regrets about leaving law school: “It’s the best thing I can do with my career and the biggest impact I can make.”