Building a Lending Vertical Inside an Edtech Company: The Three-Sided Trust Problem

The Problem That Did Not Exist Until It Did

A Nigerian student applying to a UK university can face upfront costs exceeding £30,000 — tuition deposit, accommodation, visa fees, flights, living expense proof — before stepping on a plane. They were academically qualified, had offer letters, had been accepted — and then the money problem killed the dream. We could see it in our data: a significant segment dropping off at the financial confirmation stage. Not choosing not to study. Being priced out of a system they had already been accepted into.

The Three-Sided Trust Problem

Most fintech products solve a two-sided trust problem: lender and borrower. EdFintech, in the international student lending context, is a three-sided trust problem. Every party has legitimate concerns. Every party needs something from the other two. And none of them trust each other at the start.

Student cannot get a loan without bank trust. Bank cannot assess trust without data it does not have. University will not issue CAS without financial proof. Student cannot get financial proof without a loan. A perfect circular dependency, with a real human's education on the other side of it.

“Every party in the chain was waiting for every other party to move first. Our product had to be the thing that broke the deadlock — not by removing risk, but by redistributing it in a way that made everyone comfortable enough to act.”

How We Broke the Deadlock

Step 1: Solve for the bank first

We started with banks, not students. Without bank participation, there was no product. The pitch: “we have institutional-grade data on student quality that your standard credit assessment cannot access.” Confirmed offer letters, academic profiles matched against programme completion data, visa intent signals, institutional employment track records. A fundamentally different risk signal than a credit score — and for banks active in high-sending markets like India, Nigeria, and Pakistan, it was compelling.

Step 2: Build the data infrastructure before the lending product

We spent four months building the Student Quality Framework — combining academic signals, visa intent signals, and destination employment data into an auditable, explainable risk profile. Without this structured, defensible methodology, every bank conversation ended at “interesting, but we cannot operationalise it.”

Step 3: Solve the university chicken-and-egg with a commitment letter model

Universities needed irrevocable funds. Banks would not commit without a CAS. Classic deadlock. We broke it with a staged commitment architecture: banks issued a conditional commitment letter — legally binding, contingent only on CAS issuance. Universities accepted it because it came from a regulated institution. CAS triggered disbursement. Loop broken. Six months of negotiation with the first three banking partners. Once live, every subsequent conversation moved faster.

Building the Product: What Edvoy Funds Actually Required

Simple for students. Complex underneath. Five infrastructure layers made it work:

1. 01
   Application and eligibility engine. Students needed to apply through Edvoy, not directly to the bank. We built an eligibility pre-screening tool that assessed a student's profile against the lending criteria of each banking partner before surfacing them as options. Students who were unlikely to be approved were routed to alternative funding sources rather than sent to banks to be declined. Protecting the student experience — and our conversion rate — required pre-screening at our layer.
2. 02
   Document orchestration pipeline. A student funding application required 12–18 documents: passport, offer letter, academic transcripts, visa evidence, financial statements, and more. We built a document orchestration layer that mapped each document to each bank's specific format and completeness requirements, validated documents at upload, and packaged them in the exact format each bank's underwriting team expected. What felt like a single application to the student was simultaneously a correctly-formatted submission to three or four banking partners.
3. 03
   Status and communication engine. Student lending decisions can take days to weeks. Students in this situation are anxious. Universities have CAS deadlines. Visa application windows are closing. We built a real-time status engine that gave students visibility into exactly where their application was in the banking pipeline, with proactive alerts for any action required on their part. Reducing information asymmetry for the student was as important as the lending decision itself.
4. 04
   Banking partner integration layer. Each of our 20+ banking partners had different API maturity levels, different document requirements, different decision timelines, and different regulatory environments. Some were fully API-integrated. Some required email-based submission. Some operated only in specific origin countries. We built an abstraction layer that normalised all of this complexity behind a single interface for the student, while maintaining partner-specific logic underneath.
5. 05
   Compliance and regulatory engine. EdFintech is not just regulated in one jurisdiction. A student in India applying to a UK university through a platform registered in the UK touches RBI guidelines on overseas education loans, UKVI financial evidence requirements, and the FCA's expectations around credit referral. We mapped the regulatory surface area across every student origin and destination combination we operated in, and built compliance checkpoints into the product flow at every relevant junction.

The Hardest Conversations

With banks: data sharing

Banks wanted our data but had compliance concerns about using third-party signals in credit decisions. We landed on providing a derived risk score with documented methodology — not raw student data. Banks remained decision-makers; Edvoy was an auditable data provider. This structure became the template for every subsequent partnership.

With universities: the commitment letter

The first time we presented a conditional commitment letter, the admissions office was politely sceptical — it was not on any standard list of accepted financial evidence. Three months of legal review, bank reference checks, and compliance team meetings later, three universities formally accepted it. The rest followed significantly faster.

What We Learned About Building Fintech Inside Non-Fintech

The Outcomes

The Bigger Lesson

Building a fintech vertical inside edtech sounds like scope expansion. It was scope clarification. Edvoy's job was not to connect students to universities. It was to remove every barrier between a qualified student and their education. Financial barriers were the most common reason qualified students did not enrol. The broader principle: the most defensible product expansions solve the next biggest friction in the same user journey, using assets you uniquely own. We did not build a bank. We built a bridge using data only we had. Three-sided trust problems require patience and sequencing — but when you solve them, you build something that looks, from the outside, deceptively simple. That is usually the sign the hard work was done right.