Mirari Labs

The problem

The world teaches people what to know. The defining skill of the AI age is knowing how to think.

In the conversations on AI, work, knowledge and identity are all being rewritten. The people who will thrive aren't the ones who know the most. They're the ones who can think, adapt, and reinvent themselves. We believe that's a skill that can be developed.

We started Mirari Labs because the age of AI is the greatest opportunity for human reinvention in history, if we build the right tools to meet it.

What we believe

AI should stretch how people think, not replace thinking.

We build tools that make people more capable, not more dependent. AI is at its most powerful when it helps someone understand something they couldn't before. The human stays in the loop because the human is the point.

Curiosity is the foundation everything else is built on.

Curiosity is the skill that unlocks every other skill. It's what makes people learn, adapt, and reinvent themselves. Credentials fade. The impulse to ask "why" and "what if" compounds forever.

Good ideas deserve to reach people.

Research that stays in a paper changes nothing. We take ideas from evidence to product to market, because impact is the goal. If it works, ship it. If it doesn't, learn why.

How we work

We work by co-building.

We work with a number of partners: universities, school systems, training providers, who bring the population of learners. We bring learning scientists, AI engineers and product builders, and we embed alongside their teams for the duration of the build.

From research to real education impact.

Research

We study what's emerging at the intersection of AI and learning science. Evidence first, then product. Our research shapes what we build and how we build it.

Build

We are an AI product foundry. We co-design and co-build with partners, turning bold ideas in education into working products. Our engineers, researchers, and designers ship jointly with partner teams. We ship in weeks and iterate in days.

Co-deliver

Building the tool isn't enough. We embed inside partner organisations, working in their environment, on their data, with the people who will use the tool. We stay until impact is proven, not just until launch.

Areas of focus

Agency and curiosity in learning

Students are growing up AI-native and expect autonomy in how they learn. But most education systems still run on frameworks that suppress curiosity rather than develop it. We work with schools and universities to build tools that put student agency back at the centre.

AI-durable careers

The shape of work is changing faster than the systems that prepare people for it. We partner with universities, training providers, and employers to build pathways into AI-durable careers and cultivate durable skills of the future.

Agentic AI, deployed responsibly

Agentic AI will be too hard for most institutions to build alone, yet too important to skip. We embed with partners to build agentic systems that fit their context, safe with learner data, aligned with pedagogy, and designed to augment educators rather than replace judgement.

Team

Mirari Labs was founded by Kyla Tan and Gary Liang, both former educators at the University of New South Wales. Kyla leads growth at Sana, the AI-powered enterprise learning platform acquired by Workday, was previously at Bain & Company and built social ventures to improve education at Purposeful Ventures. Gary is the founder of Bloom AI, an AI learning platform used across 20+ education institutions, and was previously at McKinsey. Together, they bring experience across AI, education, strategy, social impact and time in the classroom.

Building AI that helps humans think.