From Engineering to AI Education: Antonio Rutilio and StudiApp

I studied civil engineering and architecture at the University of Cagliari: mathematical analysis, general physics, structural mechanics — the sequence of foundational examinations every engineering student has to get through.

What stayed with me is not the content. It is how little was ever said about method. The material was taught thoroughly. How to organise six months of preparation for it was left to each student to work out alone, usually late and usually the hard way.

Two decades later, students describe the same experience. That is a strange thing for an industry to leave unsolved for twenty years, and it is why I became involved in StudiApp.

The bottleneck is real

In many engineering, science and medical faculties, a small number of foundational examinations act as a gate. Mathematical analysis, general physics, organic chemistry, structural mechanics. These are not obscure electives; they are the courses a large share of students must pass to progress at all.

Failure rates on them are high, and the consequences compound. A student who fails an exam twice does not simply try again; they fall out of sequence with the courses that depend on it, lose a semester, lose momentum, and in a meaningful number of cases leave the degree entirely. The students who drop out are frequently not the ones who could not have understood the material.

The standard explanation is that these subjects are difficult. That is true and insufficient. They are difficult in a particular way: they require the accumulation of dependent skills over months, and they punish gaps that occurred early and silently. Someone whose grasp of limits is shaky in week three will struggle with integrals in week ten, and the structure of a typical university course gives them very little chance of noticing that in between.

Why a solver is not the answer

Most educational technology aimed at this problem is built around the exercise. You are stuck, the tool shows you the solution, you move on.

It is genuinely useful, in the same narrow way a calculator is useful, and it does almost nothing for the underlying problem. Watching a correct solution appear is not the same as being able to produce one under time pressure on an unfamiliar variant. The student who works through fifty solved exercises can arrive at the exam with a completely accurate impression that they understand the material, and be wrong.

Worse, a solver has no opinion about what you should be doing. It answers the question you bring it. It does not know that the reason you keep failing integration by parts is a derivative rule you half-learned two months ago, and it does not know that the exam is in nineteen days.

The architecture of a preparation

This is where a design background turns out to be relevant, and not metaphorically.

Nobody competent begins a building by drawing. You begin with survey and analysis: what is the ground, what are the constraints, what already exists. Then you establish a programme and a plan. Then you build in a defined sequence, because some things cannot be done before others. Throughout, you verify — against the drawings, against the regulations, against reality.

Preparing for a difficult examination has precisely the same structure, and is almost never approached that way.

Survey

StudiApp starts with a diagnostic test. Not to assign a grade, but to establish the actual starting condition — which prerequisites are solid, which are approximate, which are absent. This is the equivalent of the site survey, and skipping it is why so much study effort is spent in the wrong place.

Plan

From that assessment the platform builds a personalised study plan: what to work on, in what order, with what weighting, against the time available. The sequencing matters as much as the content, because dependent skills have to be built in order.

Construction

Then the work itself — structured lessons, guided exercises and an AI tutor available at the moment a student is stuck, which is usually late at night and rarely during office hours. The tutor's job is not to hand over answers but to keep the student moving through the plan.

Verification

Finally, realistic timed exam simulations. This is the equivalent of testing the structure rather than trusting the calculation. A student who has never worked under exam conditions does not know what they can do under exam conditions, and the gap between the two is frequently the entire problem.

My role, and its limits

I participate in StudiApp as a co-founder and strategic product advisor. That means strategic direction, product development, user experience and long-term growth. It does not mean I write the mathematics content, and it would be a poor use of anyone's time if I did.

What a design and engineering background contributes is a particular instinct about structure: what the sequence should be, where a process quietly loses people, and when an interface is asking a student to make a decision they have no basis for making. These are the same judgements one makes about circulation in a building — where people get lost, where they hesitate, where a badly placed door creates a problem that no amount of finish quality will fix.

The other thing a professional STEM background contributes is a clear memory of what these subjects are actually for. Structural mechanics is not an obstacle course invented to filter students. It is the reason buildings stand up. A platform designed by people who have used the material professionally is more likely to preserve that connection, and students who can see why a topic exists learn it differently.

What measurable means here

The word that matters most in StudiApp's approach is measurable. Not because numbers are inherently virtuous, but because the alternative is a student walking into an examination with no evidence about their own readiness beyond a feeling.

A student who has taken a diagnostic, followed a plan, and sat three timed simulations knows something concrete about where they stand. That knowledge changes behaviour in the final weeks, when behaviour still matters. It is the difference between revising and guessing.

Architecture and education turn out to share the same fundamental logic: analyse the starting conditions, establish a clear plan, build progressively, and verify continuously. Applied to a building it produces something that stands up. Applied to a semester it produces a student who passes.

That is what the platform is for, and it is why an architect and civil engineer ended up helping to build it.

You can read more on the StudiApp page, or visit StudiApp directly.