Skip to main content

Systems operational

Existing client enquiries
NetEvolution
Case study / Higher education

A student-facing workflow that staff could actually run

At Royal Holloway, University of London, coursework extension requests were a manual, mailbox-driven process. A five-month engagement replaced it with an online system that students could apply through and staff could process 40% faster.

Client
Royal Holloway, University of London
Sector
Higher education
Period
2017
Role
Full stack developer (contract)

Delivered by Brad McAllister, founder of Real Code Ltd, while embedded in the institution's engineering team.

01

The challenge

Coursework extension requests are a small process with outsized load: predictable volume spikes at assessment deadlines, sensitive student circumstances behind every request, and staff rekeying details between mailboxes, spreadsheets and records systems.

It is exactly the class of problem — high-friction, rule-tolerant, emotionally sensitive — where badly designed automation makes things worse.

02

What was delivered

An ASP.NET Core web application integrated with the university's existing systems, giving students a single route to apply for extensions and staff a centralised way to review and manage them.

The work sat within a wider remit developing online systems for centralised management of courses and programmes across the Web and Mobile department.

03

Measured results

Staff processed extension applications 40% faster once requests arrived structured rather than as free-text email.

Administrative efficiency in the affected processes improved by around 30%, and delivery practices adopted during the engagement lifted project delivery speed by roughly 20%.

These figures are from the institution's own delivery reporting at the time — we quote them as measured, not extrapolated.

04

Why it matters for AI

The pattern has not changed since 2017: structure the intake, keep a human on the decision, measure the result. Agentic workflows apply the same discipline to harder problems — the difference is that agents can now handle the unstructured middle, not just the form.