Case study
How a Canadian college freed staff from the monthly payment chase
An AI agent that handles tuition outreach across email, SMS, and phone, answers financial-aid questions, and reports back what matters.
Days of admin time
reclaimed each month from manual reminder sends
3 channels
of consistent student outreach, up from email-only at scale
24/7
answers to student tuition and financial-aid questions
Overview
A Canadian private college serving several hundred post-secondary students was running its tuition collections process almost entirely by hand. Every month, an administrator pulled an overdue-students list out of the institution’s student information system, cross-checked it against payment records, and then sent individual reminder emails out one at a time. Between 50 and 250 students each month needed some form of follow-up, and the team did not have the bandwidth to consistently call or text the ones who did not respond to the first email. Inquiries about tuition balances, financial aid, and payment plans were arriving in the same inbox that the reminders were going out of, slowing everything down.
The college partnered with Squirrels.ai to deploy an AI AR Collections Specialist that connects to the SIS via API to pull updated overdue student data on a regular schedule, then runs a structured multi-channel outreach sequence across email, SMS, and phone. The agent is trained on the college’s own knowledge base for tuition policies, financial aid options, payment plans, and dispute handling, so students get accurate answers in real time instead of waiting for a reply from the bursar’s office. The administrative team now reviews a summary report each week of who has promised to pay, who has disputed, and who needs human intervention, rather than personally chasing every overdue balance.
Client at a glance
Industry
Post-secondary education (private college)
Overdue volume
50 to 250 students per month requiring tuition payment follow-up
Location
Canada
Current systems
Student information system (SIS) with API access, plus Excel as a fallback data path
Student population
Several hundred students enrolled, with active intakes across multiple programs
Use case deployed
AI AR Collections Specialist for tuition payment reminders and financial aid Q&A
The challenge
The college’s tuition collections process was reliable in the sense that it eventually got done, but it consumed enormous administrative time and left students with a slow, uneven experience.
A fully manual monthly reminder cycle
Every month, the team pulled an overdue list from the SIS, reviewed it line by line, and sent payment warning emails to each student manually. The process took days of focused administrative time, and that was just the first touch. Follow-up calls and texts were inconsistent because there simply was not enough time to do them with the same rigor across every student.
Identifying who actually needed follow-up
Pulling the overdue list itself was time-consuming. Students on approved payment plans, students with pending financial aid disbursements, and students who had already paid that morning all had to be filtered out before any outreach could happen. Errors happened, and a student occasionally received a payment warning they should not have, which damaged trust.
Volume that grew with each intake
During heavier intake cycles, the number of students needing payment follow-up could climb to 250 in a single month. At that volume, the existing manual process broke down: reminders went out late, calls did not get made at all, and the team relied on the email channel almost exclusively, which is not how a meaningful share of the student population actually prefers to communicate.
Student inquiries piling up
Even routine questions like “what is my balance?”, “can I split this into two payments?”, or “has my financial aid been applied?” needed a human to look up the answer and reply. The same administrators sending reminder emails were also answering these inquiries, creating a constant context-switch between outbound chase work and inbound support.
Additional constraints
- Student-facing communication required a careful, supportive tone, not a commercial-style debt-collection voice.
- Tuition policies, financial aid rules, and payment plan options changed across intakes and needed to be easy for the institution to update without a developer.
- Student data privacy was paramount; any new system needed clear data handling and the ability to keep production data isolated.
- The administrative team wanted summary reporting, not raw call logs, so they could focus only on the students who needed human attention.
The solution
Squirrels.ai deployed an AI AR Collections Specialist purpose-configured for a student-facing context. The college kept ownership of policy and tone; the agent took over the repetitive identification and outreach work.
Direct SIS sync with an Excel fallback
The agent connects to the college’s student information system via API and pulls a fresh list of overdue students on a regular schedule. Students with active payment plans, pending financial aid, or recently posted payments are filtered out automatically, so no one gets a reminder they should not have received. If the API is ever unavailable, the system can also accept Excel uploads as a fallback.
- Always-current contact data: the agent uses the latest student contact details and invoice data from the SIS, removing the manual data-prep step that used to consume days each month.
- Automatic exclusions: payment-plan students, students with disbursements pending, and recent payers are filtered out by rule before any outreach is initiated.
- API-first with manual safety net: Excel upload remains available for one-off lists or when the SIS is unreachable, so collections work never stops.
Multi-channel outreach with student-appropriate tone
- Email, SMS, and phone sequences: students receive a coordinated sequence across all three channels, with escalating urgency only where appropriate, never harsh.
- Supportive tone: scripts are tuned for a student audience, emphasizing options, support, and next steps rather than commercial debt-collection language.
- Auto-stop on response: outreach pauses immediately when a student pays, sets up a plan, raises a dispute, or asks for help. No one gets reminded after they have responded.
- Promise tracking: when a student says “I will pay by Friday,” the agent records the commitment and schedules a check-in instead of continuing the reminder cadence.
Dynamic FAQ knowledge base for student questions
The agent is trained on the college’s own tuition policy, financial aid rules, payment plan options, and dispute procedures. Students get accurate, on-brand answers in real time on the same channel they were contacted on.
- Tuition and balance questions: the agent confirms the amount owed and explains what the balance covers.
- Financial aid status: students can ask about disbursement timing and pending aid, and the agent provides a consistent answer pulled from the knowledge base.
- Payment plan options: when a student asks for a split or extension, the agent explains the available plans and flags the request for staff review.
- Disputes flagged clearly: any student who challenges a charge is removed from the cadence and escalated to a human reviewer with the full conversation history attached.
- Easy updates: the college updates the knowledge base directly whenever policies change. No engineering, no waiting.
Summary reporting, not raw call logs
- Weekly digest of who promised to pay and by when, who is in dispute, and who still needs human follow-up.
- Channel-level analytics on which mix of email, SMS, and call is producing the best response rates.
- Full conversation transcripts are available on demand when an administrator needs to dig into a specific student’s history.
- Administrators spend their time only on the students who actually need them, instead of triaging the entire overdue list.
Privacy and data protection
- Strict data isolation between customer environments, with NDAs and secure cloud protocols in place.
- Student records are never used outside the college’s own context.
- Outreach automatically respects do-not-contact flags so any student preference is honored on every channel.
Results and impact
The college moved from a fully manual monthly reminder cycle to a continuous, multi-channel collections motion that the administrative team supervises rather than executes.
What changed in the business
- Manual reminder work effectively disappeared. Days of monthly admin time spent pulling lists and sending reminder emails one by one are now handled automatically, with the team reviewing summaries instead.
- Every overdue student gets multi-channel follow-up. What used to be an email-only motion at scale is now a coordinated email, SMS, and phone sequence applied consistently across all 50 to 250 students each month.
- Students get answers right away. Tuition balance, payment plan, and financial-aid questions are answered in real time from the knowledge base, instead of routing through the bursar’s office inbox.
- Mistakes from manual filtering disappeared. Students on payment plans, pending financial aid, or recent payments are excluded by rule, so no one receives a reminder they should not have.
- Staff focus shifted to the cases that matter. Administrators now spend their time on disputes, hardship cases, and payment-plan setup, rather than triaging an entire monthly list.
- Student-appropriate tone, consistently applied. Every conversation follows the same supportive, on-brand voice, no matter which channel or which student.
- Scales with intake. When the overdue list spikes to 250 students in a heavy intake month, the system handles it without needing extra hours from the team.
TESTIMONIAL
Why Squirrels.ai
Squirrels.ai deploys AI agents that handle calls, emails, and texts at scale, trained on a company’s own scripts, FAQs, and SOPs. For this college, the platform delivered:
- An AI AR Collections Specialist purpose-tuned for a student-facing context, with a supportive tone instead of a commercial collections voice.
- Direct SIS API integration with automatic exclusions for payment plans, pending aid, and recent payments, plus an Excel fallback for resilience.
- A dynamic knowledge base for tuition, financial aid, payment plan, and dispute questions that the college can update directly any time policies change.
- Summary reporting that highlights only the students who need human attention, instead of dumping raw logs on administrators.
- Strict data isolation, NDAs, and secure cloud handling, with respect for student do-not-contact preferences across all channels.