Online education only works when students feel seen, supported, and connected. Our work with Risely showed that responsible integration of AI can help us scale human care, not replace it. It allowed our team to engage students, strengthening the kinds of relationships and interventions that drive real student success.
How the University of Arizona 32x'd proactive student outreach in 8 weeks.
Arizona Online is the fully online division of the University of Arizona, serving roughly 11,000 students pursuing degrees and credentials from the University's flagship campus. At the center of that experience sits a ten-person Student Engagement Specialist (SES) team, responsible for supporting 7,300 online undergraduates as they navigate the institution and the resources around it.
In February 2026, Arizona Online set out to test a specific hypothesis: that a unified, AI-assisted platform could give a ten-person team the leverage to proactively support 7,300 undergraduates. They partnered with Risely to run an eight-week pilot.
A high-performing team ready to operate at a new scale.
Online students experience the university entirely through digital channels, which means disengagement can be silent, and the students most in need of a check-in are often the least likely to initiate one.
The Arizona Online SES team had built a strong engagement practice grounded in deep care for students. The question was how to extend that practice across a caseload ratio of 1:730, ten specialists supporting 7,300 undergraduates, at a scale no team in higher ed had achieved before. The data that could surface which students were beginning to slip lived across three institutional systems - the SIS, the LMS, and the CRM - and a November 2025 workflow survey quantified the manual lift required to work across them at roughly 190 hours per week across the team. In Spring 2025, the team tracked 350 proactive emails across the 16-week term.
The pilot was designed to answer one question: could a unified, agentic AI platform give a ten-person team the leverage to proactively support 7,300 undergraduates, reaching students at the moments that matter most across their journey?
The work became more strategic, not just faster.
Risely unified data from the SIS, the LMS, and the CRM into a single triaged caseload view. For the first time, a Student Engagement Specialist could open one screen and see exactly which of their 730 students needed attention today, and why, with the data already assembled, the briefing already drafted, and a personalized outreach ready for review.
A Specialist's Day
Deep expertise, applied across a complex data environment. Context-gathering across the SIS, LMS, and CRM was a real share of the day.
- 1
Open the CRM
An unsorted task list. No way to see which students needed attention first.
- 2
Pick a student manually
No risk score, no synthesized context. Prioritization rested entirely on memory and intuition.
- 3
Hop between systems for context
Holds in the SIS, course activity in the LMS, history in the CRM, each in a separate tab.
- 4
Draft and send outreach
Personalized emails written from scratch. No way to scale that personalization across a 730-student caseload.
- 5
Return to the CRM, log by hand
Mark the task complete, write the note, move to the next.
A Specialist's Day
The same judgment, freed from the manual overhead. Specialists open the platform to a caseload already triaged, briefings already drafted, and a queue ready for review.
- 1
Open Risely
A unified workspace. SIS, LMS, and CRM signals on one screen.
- 2
Caseload already triaged
Holds, course activity, and recent interactions weighted overnight.
- 3
Briefings already drafted
Records, case notes, and milestones synthesized into a quick pre-session summary.
- 4
Personalized outreach drafted
Grounded in the student's actual data. The specialist reviews, edits, and sends.
- 5
Conversations captured automatically
Calls and meetings transcribed and synced back to the CRM.
With the manual overhead handled by the platform, the team brought their practice to a scale that hadn't previously been possible. They designed campaigns around specific populations: students with enrollment-blocking holds, students who hadn't yet enrolled in the next term, first-term students entering mentorship, students making strong progress who deserved a congratulations. They surfaced institutional patterns across their caseloads, hold categorization issues, transfer-credit mismatches, that no individual student would have raised on their own. And they reinvested hours into the 1:1 conversations that needed a human: phone calls, video meetings, extended email threads with students whose situations didn't fit a standard script.
A second-order effect emerged partway through the pilot. The Arizona Online team's depth of operational knowledge meant they quickly began shaping the platform itself, proposing new filters, new campaign types, and new use cases the original scope hadn't anticipated. Risely shipped against those requests continuously. The more the team engaged with the platform, the more ideas they generated, and the more the platform came to reflect the way Arizona Online actually works.
The human-in-the-loop model was important to our team. Every automated message was reviewed, edited, and sent by a student engagement specialist. AI didn't replace that judgment; it removed the manual work that was getting in the way of it. Coaches went from hunting through multiple systems to gather information to a coordinated student briefing that brought all of that information together in one clean overview, giving them the time back to spend with students who need the support.
From 22 to 692 proactive outreaches per week.
The pilot's results reflect the operational rigor of Caleb Simmons, Jessica Salata, and the ten-person SES team, whose deep knowledge of online student needs shaped every iteration of the platform.
In eight weeks, the SES team created roughly 1,400 new hours of proactive-engagement capacity, 41 work weeks of added output, the equivalent of six additional full-time specialists, without a single new hire. The 10-person team is now producing the proactive outreach volume of a team of roughly 16.
The leading indicators of retention all moved in the right direction.
An eight-week window is too short to measure retention, progression, or re-enrollment, and the team has been deliberate not to claim otherwise. What the pilot does demonstrate is that the leading indicators of retention, whether at-risk students are being identified, whether they are being reached, and whether they are responding, are all moving meaningfully in the right direction.
The students who responded weren't the students already filling the queue. They were the ones who had been invisible, and the themes their replies surfaced are what proactive outreach makes visible at scale for the first time.
Time absorbed by the platform was reinvested where the team's judgment matters most: creative problem-solving, strategic campaign design, data-driven intervention, and real 1:1 human relationships.
The real impact came from pairing technological innovation with a team deeply committed to student success. By reducing the manual barriers that often consume student support work, our specialists were able to spend more time building meaningful relationships, identifying challenges earlier, and engaging students with the kind of care and intentionality that helps them persist and thrive.