👤 Allen Bell (Agency Director, Midland County CSCD)
Allen Bell attended the University of Texas at San Antonio, where he participated in Army ROTC and graduated with a BA in Criminal Justice. Immediately upon graduation, Allen was commissioned as a second lieutenant in the United States Army Military Police Corps and served in Saudi Arabia, Kuwait, and Iraq during Operation Desert Shield/Desert Storm. Allen began working as an adult probation officer with the Midland Judicial District CSCD in 1991, where he is still employed. He earned a Master of Science degree in Criminal Justice Management from Sam Houston State University in Huntsville, Texas, in 1995. He is currently the Director of the Midland CSCD, a position he has held since 2015.
This guest post is based on a presentation Director Bell provided at the APPA Training Institute Conference in New York, NY in August 2025.
When I told my kids I was going to write a guest post on artificial intelligence (AI), they laughed and reminded me that they used to program the VCR for me. But here’s the beauty of AI: you don’t need to be a technical genius to use it. Just like setting the clock on a microwave, it’s accessible to everyone.
In this article, I want to share how our agency - the Midland Judicial District - has been able to leverage AI-based technology to empower our officers and enhance client outcomes. Over the past several months, we’ve piloted an AI-driven platform called Atlas from The Change Companies. Our journey highlights both the challenges we face in community supervision and the ways this technology can support staff and improve client outcomes.
As an administrator, my goals are simple: to drive better outcomes and reduce re-offending. But the reality on the ground is tough.
Our probation officers and counselors carry high caseloads. They want to build meaningful connections with clients, but most check-ins are brief, surface-level conversations: “How are you doing?” “I’m fine.” That lack of depth is a missed opportunity to intervene before problems escalate.
Generic, one-size-fits-all approaches don’t work either. That’s why we conduct risk and need assessments—but tailoring services to each individual is easier said than done.
Why we chose Atlas
To address these challenges, we launched a pilot in April within our triage unit, a specialized counseling program staffed by licensed chemical dependency counselors and a recovery support peer specialist. Clients in this unit are identified using the Texas Risk Assessment System (TRAS), which pinpoints both risk level and need areas.
Here’s how Atlas fits in:
- Tailored Journals – Based on assessed criminogenic needs, clients are assigned digital Journals. We use the Courage to Change - a risk-need-responsive community supervision system that supports effective case management.
- Flexible access – Clients complete Journals on their phones, tablets, or computers. For many in West Texas, especially those working long shifts in the oil fields, this flexibility is critical. They can complete Journals at 1 a.m. on a rig or during downtime, turning unstructured hours into prosocial time.
- AI-generated insights – This is where Atlas transforms the process. After clients complete their entries, the AI analyzes the content and generates summaries, highlights key themes, and even recommends follow-up questions. Staff no longer have to sift through pages of entries before each session. Instead, they can immediately see where a client is in their change process.
Voices from the field
One of my program coordinators, a licensed counselor leading this effort, explained how Atlas has changed her work:
“The summaries and insights save so much time. I’m not putting words in a client’s mouth—it’s their own journaling, reflected back. That lets me focus on asking the right questions instead of wondering what to do next.”
We also brought in Personal Development Seminars, a Texas-based group that facilitates classes via Zoom. Their CEO, Max Maurer, shared:
“I can catch red flags in clients’ journaling before a session even begins. The progress summary tool highlights both areas to encourage and areas of concern. It lets us address warning signs early—before a relapse shows up in a UA or an arrest.”
From reactive to proactive
Traditionally, supervision is reactive: we wait for a failed drug test or a new arrest before acting. Atlas shifts that dynamic. The system flags language suggesting frustration, cravings, or criminal thinking, alerting staff to potential risks. It also recommends conversation starters and even next Journal assignments. This helps staff plan sessions more effectively and intervene earlier.
Early results
Our pilot has shown promising outcomes:
- 84% engagement rate – far higher than with paper Journals, which were often lost or forgotten.
- 10 hours of staff time saved per week – thanks to AI-generated summaries.
- Increased voluntary participation – clients are starting additional Journals on their own, finding value in the process.
Just as importantly, we’re seeing a shift in mindset. Clients are more engaged, and staff feel supported rather than burdened.
Final thoughts
Let me be clear: this technology is not a replacement for probation officers or counselors. It’s a tool that helps them do their jobs better and more efficiently.
Younger staff in particular expect to use technology at work—it’s the world they’ve grown up in. If we fail to embrace these tools, we risk falling behind, both in effectiveness and in retaining our workforce.
For us in Midland County, integrating AI into supervision has been less about replacing human judgment and more about enhancing it. By combining professional expertise with AI-driven insights, we can move closer to our ultimate goal: improving lives and reducing recidivism.
