Applied AI NL
case studies · portfolio of implementations

No theses. Working solutions.

Every student spends four years building AI solutions in their own organisation — this is the kind of work that produces.

Anonymised, representative examples of student projects.

Scheduling intake in healthcare

Healthcare organisation · year 3 · Gen AI Makers

Problem

Scheduling changes arrived by email and phone with a single planner. Requests piled up, priorities were unclear and teams waited days for a decision.

Approach

The student first mapped the entire request process: who requests what, through which channel, with what urgency. She then designed an intake bot with an agent-workflow behind it, within the Gen AI Makers module.

Built

An intake bot that gathers requests in a structured way, plus an agent-workflow that prioritises them and stages them as proposals in the scheduling system. The planner approves each proposal — the human retains the final decision.

Result

The turnaround time for scheduling changes went from days to hours. The planner now has time for the exceptions that genuinely require human judgement.

A single source of truth for customer data

Logistics SME · year 2

Problem

Customer data sat polluted and scattered across three systems: differing addresses, duplicate records, outdated rates. Quotations regularly went out based on incorrect data.

Approach

The student organised data quality according to DMBOK principles: unambiguous definitions for each customer attribute, explicit ownership per system and measurable quality rules. No big bang, but a manageable set-up.

Built

A set of documented definitions and measurement rules, designated data owners and a monthly quality report that makes discrepancies between the three systems visible.

Result

The error rate in quotations demonstrably decreased. There is now one agreed source of truth for customer data — and a report that ensures it stays that way.

Decision support for subsidy applications

Municipality · year 4 · graduation project, 20 weeks

Problem

The assessment of subsidy applications was slow and inconsistent: different assessors interpreted the same policy rules differently, and applicants waited weeks for a decision.

Approach

In a 20-week graduation project, the student designed RAG-supported decision support based on the organisation's own policy documents, applying the frameworks from Reliable Decision Making & Compliance Monitoring: explainable, logged and with the human as final decision-maker.

Built

A RAG application that retrieves the relevant policy passages for each application and drafts a substantiated preliminary recommendation. Every step is recorded in compliance logging; the assessor takes the final decision.

Result

Assessments are more consistent and turnaround times are shorter. Every recommendation is fully auditable and traceable to policy text and assessor.

Monthly reporting without manual spreadsheet work

Financial services firm · years 3–4

Problem

Monthly reports were compiled manually from spreadsheets supplied by different departments. This cost the controller days every month, and errors crept in easily.

Approach

The student first designed a data model that brings the departments' source data onto a common footing. On top of that, an agent team with a clear division of tasks: collect, check, draft — with the controller as client.

Built

A data model plus an agent team that collects source data, checks it for consistency and drafts the preliminary report. The controller approves the draft and enriches it with interpretation and context.

Result

The report is ready days earlier. The controller shifts from producing to analysing — from retyping figures to explaining them.

Your portfolio

You will build this portfolio too

From year 1 onwards you work on real assignments in your own organisation — not practice cases, but challenges that matter to your colleagues. After four years you have a portfolio of working implementations: proof of what you can do, for your current and your next employer.

what it does for your career → what it does for your employer →
Curious what you would build?

Discuss your own case

Schedule a short call, or email or phone the programme.

Schedule a 15-minute call ict-deeltijdopleiding@hu.nl 06 82 094 452