HELP University students have completed AI modules that adjust in real time to individual performance, culminating in live hackathon-style “Sprint Finals” judged by industry panels — with the model now expanding into a corporate AI training track under a signed MOU.

Kuala Lumpur, August 5, 2026 — LAPA Labs, a Malaysian AI EdTech startup, and HELP University today outlined results from their pilot of an adaptive AI curriculum now running across HELP’s IT undergraduate programmes, alongside a newly signed MOU extending the same model into an HRD Corp (HRDF)-claimable corporate training track.

Unlike fixed course sequences, the LAPA platform routes each student to their next challenge based on three real-time signals — response time, iteration count, and output quality — rather than a syllabus set months in advance. Students who clear a challenge quickly move to harder, more ambiguous problems; students who struggle get another pass with more scaffolding before advancing.

Students who have completed a module report a 100% module completion rate, with an average 70% capability score on final assignments. The most recent Live Sprint Final, run jointly with Disted College and focused on homelessness, urban poverty, and mental health in Malaysia, drew 4 participating teams and over 1,200 public votes on the winning solutions.

Why It Matters

Malaysia’s Digital Ministry estimates AI could add RM13–20 billion a year to GDP by 2030. Malaysian public universities have rolled out AI tools to hundreds of thousands of students this year, across a higher education system enrolling more than 1.2 million students — a scale at which a single fixed curriculum leaves students at very different starting points learning at the same pace.

“If we teach AI using only textbooks, our curriculum risks obsolescence before the semester even ends,” said R Murali Rajaratenam, Senior Lecturer and academic lead for the project at HELP University. “We needed our students actually building things with AI, not just learning about it, the same way they will the day they start working.”

How The Model Works

Each lesson feeds into a project scored on a running scoreboard, building toward a Live Sprint Final: a real problem, a tight build window, and a pitch to independent judges — run as a hackathon rather than a classroom presentation. In the most recent Sprint Final, teams had 80 minutes to build a solution before a 10-minute pitch to judges, with cash prizes awarded to the winning team.

An AI grading layer scores student output on the same signals used for routing, paired with human faculty review — HELP’s lecturers set what “good” looks like in each discipline before any lesson is built; the AI layer speeds up feedback delivery, not the judgment behind it. Every assignment returns specific, individualised notes on what worked and what to improve next, delivered weekly.

The system is cloud-hosted and API-first, with model credentials kept server-side. The same backend now also runs HELP’s HRD Corp (HRDF)-claimable corporate track.

“Our lecturers set the direction, what needs to be taught, what good looks like in the field,” said Firdaus Salim, co-founder of LAPA Labs. “We build the practice environment around it, fas

Origins in Competitive Esports

LAPA’s pedagogy draws from competitive esports rather than traditional EdTech: immediate stakes, instant feedback. “Traditional EdTech platforms are essentially digital filing cabinets — they don’t drive imagining or building,” Firdaus said. “We built LAPA to bring the mechanics of a live match to AI skills training.”

For HELP student Harenthirah Gopala Krishnan, a Bachelor of Computer Science student in the pilot, the effect is a course that doesn’t feel paced for an average student he doesn’t resemble. “When that daily challenge drops and the timer starts, you have to think fast and actually figure out how to make the AI do the heavy lifting for you,” he said. “It doesn’t feel like a class, it feels like you’re actually doing the job you’re studying for.”

What’s Next

LAPA and HELP University have formally signed an MOU launching an HRD Corp (HRDF)-claimable corporate track using the same adaptive model, targeting 20 Malaysian companies and upwards of 500 employees across finance, manufacturing, and technology sectors in its first phase. LAPA is also in early discussions with institutions in Singapore, the Philippines, and Vietnam.

“We want students to be in awe of what they can create, and how they can make workplaces, communities, and the world better with it,” said Firdaus. “We start by making sure no student is stuck on someone else’s timeline.”

This press release has also been published on VRITIMES