Financial Services Firm Reduces Cost-to-Serve with Agentforce
An FDE squad deployed governed Agentforce service agents with escalation and observability, letting AI absorb repetitive volume while humans focus on complex work.
Read case studyA 1,240-student coaching institute was losing 15 to 20 percent of its admin bandwidth to manual fee work: rebuilt spreadsheets every cycle and hundreds of one-by-one WhatsApp reminders. With Feezy, the entire fee operation was live and automated in a single onboarding session, and the first automated cycle landed a 94 percent collection rate.
The client is an established multi-subject coaching institute in India serving 1,240 active students across JEE, NEET, and foundation programs, spread over several batches, courses, and academic years. It is exactly the kind of business that defines the segment: profitable, locally trusted, and run by a small team where the founder still teaches and the same one or two admins hold the entire operation together.
That profile is also why generic education ERPs had never stuck. The institute had the volume to make manual fee work genuinely painful, but not the dedicated IT staff or appetite for a six-figure implementation that enterprise systems assume. So, like most of its peers, it ran its core revenue process on a spreadsheet and a phone.
Industry
Education: coaching institute
Size
1,240 active students across multiple batches
Geography
India
Product
Feezy: fee collection and management
Before Feezy, every fee cycle was a project in its own right. The institute estimated it was spending 15 to 20 percent of its admin bandwidth on fee management alone, work that should take minutes a month, stretched into days. Three failures compounded each cycle.
Schedules rebuilt from scratch, every cycle. Fee schedules, installment plans, late fines, and discounts were re-entered by hand each term because the previous sheet was a tangle of merged cells and one-off formulas that no longer matched the current batches. The work added no value, and it was fragile: one dragged formula or wrong row and the numbers went silently wrong, usually discovered only when a parent disputed an amount.
Collection chased one WhatsApp message at a time. An admin opened each chat, checked whether that family had paid, typed a reminder, attached a payment link, and moved on, multiplied across hundreds of families. There was no single view of who had paid, who owed, and who was simply ghosting. The same parent got reminded twice or not at all. The "system" was the admin's short-term memory.
One person held everything. The fee logic, the informal "pay next week" promises, and the running mental tally all lived with a single admin. There was no audit trail and no handover. When that person took leave, the institute did not lose a file; it lost its financial memory. And no one could answer the only question that mattered in real time: how much have we collected, and how much is still owed?
We did not run a months-long implementation. The engagement followed Feezy's standard onboarding playbook: a guided, self-serve setup designed to take a coaching institute from signup to a live, automated fee system in a single sitting. The principle is simple: import the data once, configure the rules once, and let every cycle thereafter run itself.
The session walked the owner and lead admin through three deliberate moves, in order, so that nothing downstream had to be redone:
The goal was never a better spreadsheet. It was to remove the manual cycle entirely, so the work that consumed a fifth of the team's bandwidth simply stops being a human job.
By the end of the onboarding session, the institute had a complete, automated fee operation running on Feezy. The pieces map directly onto the failures they came in with.
Set once, runs forever. Every cycle now generates invoices automatically from the configured templates and installment schedules. Fines apply on their own, waivers are honored, and receipts send themselves, with zero manual input after setup. The spreadsheet rebuild disappeared entirely.
Instead of an admin typing reminders chat by chat, the AI reminder engine learns when each parent actually pays and fires WhatsApp reminders at the optimal moment, not just a blanket blast on day one. The one-by-one chasing became zero-touch.
The institute got its own branded payment portal with OTP login and instant e-receipts: a premium parent experience that replaced ad-hoc payment links and screenshots of UPI IDs. Every payment is captured, attributed, and receipted automatically.
A live collection dashboard replaced month-end guesswork. The owner can now answer "where do we stand right now?" instantly, with role-based access so the fee manager, teachers, and parents each see exactly what they need and nothing more. The defaulter predictor flags students likely to pay late before the due date, so the team can act early rather than chase after the fact.
The change was measured in the very first automated cycle. The before-and-after is stark: an admin who used to spend days each term rebuilding sheets and typing reminders now spends minutes reviewing a dashboard.
Beyond the numbers, the experience changed for the people who run the institute. The owner stopped reconstructing the financial picture in their head and started reading it off a screen. The single-admin risk that used to threaten the whole operation dissolved into a system with a full audit trail and a clean handover. And the parents got a calmer, more professional experience: a branded page, an instant receipt, and reminders that arrived when they were actually useful rather than as a barrage on day one.
The work was not optimised; it was automated away. The fee operation that used to consume a fifth of the institute's bandwidth now runs on its own, cycle after cycle. As the Feezy team puts it, the fees collect themselves.
This is the part owners find hardest to believe until they see it. There was no phased rollout and no multi-week change program.
"For years, fee collection ate our evenings. We rebuilt the same spreadsheet every cycle and chased parents one message at a time. We went live on Feezy in a single afternoon, and the next cycle just collected itself: 94 percent, without a single reminder typed by hand. I finally know exactly where we stand at any moment."
Owner, 1,240-student coaching institute (name withheld on request)
What changed was not the discipline of the team, which was already disciplined. What changed was the model: a permanent student identity, automated cycles, AI-timed reminders, branded payments, and a real-time view of collection health, all running on a platform built specifically for how a coaching institute actually works. The result is the outcome this case study set out to prove: Excel to fully automated fee operations, in under two hours.
Other outcome-focused engagements from ConvertEdge Tech.