LA Public Defender CIO digitizes to divert people to programs, not prison

Mohammed Al Rawi was appointed as the first CIO of Los Angeles County Public Defender’s (LACPD) office roughly five years ago, signaling the beginning of an era where technology and justice intersect to help the most vulnerable in the court system.

He arrived at a crucial time when the office, laden with paper records and outdated legacy systems, desperately needed a digital overhaul. So he was recruited to create a 20-year vision to establish a client-centered case management system to replace the “on-premise computer closets that was called a data center,” he says.

Therefore, his first order of business was to do what was necessary to migrate data and legacy practices to the cloud, and develop a client-centered case management system to increase efficiency. So he transitioned the public defenders from exclusively using desktops to employing laptops and mobile devices. Then, he decommissioned all 24 of the office’s legacy, in-house systems and migrated them to cloud platforms, including Salesforce CRM, Microsoft Azure, Office 365, AWS, and Box.com, enabling employees to securely access information from any device from anywhere. Fortuitously, this was implemented just before the pandemic enforced stay at home orders. 

“We were prepared,” Al Rawi says about the first major phase of the digital transformation, where public defenders in the nation’s largest county could access digitized data and collaborate across 60 courtrooms.

Aside from the technology migration, Al Rawi and his team customized their mission to “humanize” the indigent, not just digitize caseloads, by leveraging Salesforce and AI to champion the rights of clients facing charges in a system that too often views them as case numbers rather than individuals.

Al Rawi shares the passion of the public defenders’ office and employs his digital talents to help divert people to “programs, not cages,” he says, paraphrasing his boss’s key objective to ‘get people out of jail and keep them out of jail.’

“Our job is to protect the Sixth Amendment and provide zealous representation to clients who cannot afford a private attorney,” says Al Rawi. “There’s a term called indigent defense, but when it comes to criminal courts and long trials, everyone is indigent. No one can access trial lawyers unless you have half a million in the bank. Everyone needs a public defender.”

Time well spent

The technology used in the past 50 years functioned like “glorified calendars,” he says, and focused only on the case and the crime in a way that dehumanized people who make mistakes — many unintentional, or committed by those with special needs.

“Everyone in the courtroom knows about what this client did, but they don’t know the client,” Al Rawi says. “Our job is to tell the story, to humanize that client at the worst moment in their lives,” adding that even misdemeanor crimes can derail a person’s life. He points to working clients who sometimes fail to appear at court because they forget the date and are arrested, lose their job, and sometimes end up homeless.

In total, it took the CIO’s team and agency a little over two years to convert 160 million documents into a transformed, revamped, and people-centric system, built on the Salesforce CRM, that tells their stories and focuses on people outcomes, not case outcomes.

Noah Cox is a deputy public defender in the agency’s neurocognitive disorders team, whose representation of a client in a well-known documentary named Forgiving Johnny underscores the disadvantages underfunded public defenders have historically faced poring over mountains of paper to represent clients unfairly caught up in the criminal system. Access to digitized records as well as analytics and AI tools gives public defenders like Cox the time required to present clients’ cases more thoroughly and build a better defense, which in Johnny’s case led to a treatment program instead of prison time.

Cox appreciates how invaluable the CIO’s vision and digital transformation has meant to the office’s mission.

“Mohammed Al Rawi and his amazing team of IT professionals have harnessed the strengths of technological innovation for our advocacy, thereby facilitating quicker assessments of our client’s needs, improving evaluations of their legal options, and ultimately paving a more efficient and effective path to rehabilitation and justice,” says Cox, whose office handled roughly 140,000 cases in 2023. “Absolutely, the digital transformation has been instrumental in our work and has contributed to changing countless lives.”

The developing client-centered system

Al Rawi is proud of the office’s use of the cloud and the creation of what might be the first indigent defense data lake ever established, on Azure. AWS and Microsoft cloud providers help him employ Machine learning (ML) and AI models to identify documents, transcribe videos and 911 calls, and use Microsoft Cognitive Services, for instance, to enable facial recognition data inclusion in cases. He deployed automated text message systems and tools such as Twilio as well to proactively remind clients when to show up for court.

ML models are also used to detect relevant documents coming in from 99 different arresting agencies, as well as efforts in gen AI, with OpenAI and Salesforce, to achieve a digital maturity that gives public defenders the best data they need to represent clients.

One ML model the IT team developed, for example, actually learns the pattern of the case summary, says Al Rawi, and provides the rap sheet, arresting officer, criminal history record, all automatically extracted and organized for the public defender, who just hits submit to get the document summarization, saving them up to 30 minutes in data entry.

Another model transcribes video into natural language processing to save time for defenders. “Video is another part of that tsunami of digital evidence we’re getting now,” he says. “We’re averaging 12 hours of video footage per case, and attorneys who have between 20 and 30 cases don’t have time to watch all that video.”

Yet another model may detect a probation violation from video, and proactively remind a client with a history of violations to make up a session they may have missed and are unaware of it. “It’s all to help the defense case and help the client get a favorable outcome,” Al Rawi says, noting the proactive use of this data prevents clients from sitting in county jail for weeks — or worse. “I’ve seen many clients who got exonerated who are innocent but could have languished in state prisons for years.”

He recalls one case in which a client was in the wrong place at the wrong time, but the data made available to the public defender helped him be ahead of the curve and prove his innocence of a murder charge. “The guy gave me a big bear hug in tears and said because of you, I can enjoy my favorite tacos with my mom every Sunday,” he says. “There’s no other job in the world that has that type of ROI, that sense of accomplishment.”

Gen AI’s potential to prevent more cases

All this time saving not only helps public defenders build stronger cases but persuades judges to send clients to diversion programs, rehabilitation programs, and treatment programs as alternatives to incarceration, which presents funding challenges but averts the costs of prison time. “Attorneys brute force check any opportunity to help them get back to their families and re-enter their communities safely,” Al Rawi says.

The LACPD has a lot of data, and Al Rawi envisions using the power of gen AI to weave patterns together and create a proactive tool to get help for the client before they become a client.

“The key is allowing public defenders to have time to focus on the legal aspects and best administrative aspects of a case proceeding, and use AI to reduce the amount of document review, document generation, document categorization, and indexing information so it’s more accessible and allows a far more contextual search than a keyword search,” Al Rawi says. “Generative AI is going to significantly reduce the mundane, very tedious business processes and automate them in a very accurate and dependable way to help lawyers have more time to help people.”