Every major technological shift presents a fundamental choice: do we use new tools to standardize and categorize people, or to understand and support them? Artificial intelligence is no exception.
Too often, educational software reduces human potential to test scores and rigid rubrics. We take the opposite stance: technology should help us notice what is already there — the curiosity, practical ingenuity, and quiet strengths that standardized assessments routinely overlook.
Every day, learners demonstrate real abilities that never make it into an official certificate. Closing that gap — making genuine competence visible without reducing people to numbers — is why grow.file exists.
Decades of developmental psychology and educational science tell us how learning happens and how diverse abilities develop. Yet very little of that insight reaches the moments where it matters most: a busy morning in a Kita, a packed classroom, or a career counseling session.
grow.file connects scientific research directly with everyday practice. Rather than building another administrative form or a black-box scoring algorithm, we build tools that help teachers, educators, and mentors notice and reflect on real moments of growth.
Research and engineering work in the same system here, led by domain experts in both fields. Together, they create practical tools that neither discipline could build in isolation.
With every feature we design, our first question is simple: Does this help someone understand their own strengths better, and help the people supporting them provide meaningful encouragement?
The skills that matter most in modern life and work — critical thinking, collaboration, resilience, self-organization — rarely fit into traditional multiple-choice exams. While the world of work is rapidly evolving, our systems for evaluating people are still largely built on methods from the last century.
When someone's practical abilities go unrecorded, they lose the opportunity to build on them, advocate for themselves, or find the right path forward.
Making competence visible is a question of fairness: ensuring that every person's real contributions and capabilities count, and that no one is overlooked simply because they don't fit into standard testing boxes.
AI is not free: every inference we run draws on data centres that consume electricity and water. We would rather be honest about that than quiet about it — and then do more than the minimum in response.
Helping a person see what they are capable of, and making sure there is a livable planet for them to be capable in, are the same goal over different timescales.
We calculate the CO₂ footprint of the inference we run and offset it through verified climate projects.
AI inference uses significant water for cooling. We account for it and offset accordingly.
We don't just neutralise our impact — we double it. For every unit of harm, two units of restoration.
What we help people become shapes the adults they grow into. The planet we leave them shapes everything else. We refuse to choose between the two.