Prantara Sign in
Our values

An open landscape of learning.

Machines can now produce the answers. The education that matters is the one that changes what a person can understand and do on their own. Prantara is built for that education, and for every learner who wants it.

Why now

AI can write the code, solve the problem set and draft the report. That makes it easy to look capable without becoming capable. It also makes something new possible: patient, individual teaching for everyone, the kind that once only a great personal tutor could give.

Which of these futures a learner gets depends on design, not on the technology. In a randomized trial with about 1,000 students, open access to a chatbot raised practice scores by 48% but lowered scores on the exam taken without it by 17%; a version that gave hints instead of answers removed the harm.1 In a Harvard physics course, a tutor built on teaching principles, with expert solutions in hand and answers held back, helped students learn more, in less time, than an active-learning class.2

Five commitments

  1. Struggle first, then help

    Understanding is built when you produce something: a set-up, a first step, a prediction. Reading a finished answer feels like learning and mostly isn't.

    On Prantara the tutor asks what you've tried and gives hints one at a time. Worked solutions open after you've made an attempt.

    Across 53 studies, working on a problem before being taught how beat being taught first.3

  2. Measure what you can do on your own

    Practise with AI as much as you like. What counts is what you can do without it, because that's what you carry into the exam room, the lab and the job.

    On Prantara every skill has two scores: one for all practice, one for work done on your own. Work done without hints or the tutor, and AI-free quizzes, show you the second.

    Learners misjudge their own learning: in the chatbot trial they didn't notice the harm,1 and in a physics class they learned more from active learning while feeling they learned less.4

  3. Return, and go deeper

    Ideas stick when you come back to them: recalled from memory, after a gap, mixed with others, a little harder each time. That is the spiral in our logo.

    On Prantara skills come back for review on a schedule, and practice mixes topics the way exams do.

    Retrieval practice and spacing are among the most reliable findings in learning science;56 in engineering, spaced retrieval improved exam scores and retention into the next semester.7

  4. Explain it to own it

    If you can explain an idea and defend your own work, you understand it. If you can't, you've found exactly what to study next.

    On Prantara you teach concepts to an AI classmate who asks hard questions, and you defend your own solutions in short oral checks.

    Explaining to yourself while you learn has a reliable, sizeable effect on what you learn.8

  5. People stay in charge of judgement

    AI serves the relationship between a teacher and a learner; it doesn't replace it. Teachers decide what is taught and what counts, and a person is always within reach.

    On Prantara instructors see where the class is struggling, steer the tutor, and review or overrule its scores. “Ask a person” sends any question to the course staff.

    This one is a commitment, not a finding.

The name and the spiral

Prantara /prænˈtɑːrə/, said “pran-TAH-ruh”, joins two Sanskrit roots: pra, “forward”, and antara, “within; the space between”. Learning happens in that space, between being shown how and being able to do it yourself.

The logo is a spiral grown from a seed. Understanding starts within the learner, and it moves forward by coming back around to the same ideas, each turn a little wider. The open end is where you go next.

We build Prantara course by course, with the people who teach them, and we hold ourselves to the standard we hold learners to: what counts is what you can do on your own.

Create an account Sign in

Sources

  1. Bastani, H. et al. (2025). Generative AI without guardrails can harm learning: evidence from high school mathematics. PNAS 122(26). doi:10.1073/pnas.2422633122
  2. Kestin, G. et al. (2025). AI tutoring outperforms in-class active learning: an RCT introducing a novel research-based design in an authentic educational setting. Scientific Reports 15:17458. doi:10.1038/s41598-025-97652-6
  3. Sinha, T. & Kapur, M. (2021). When problem solving followed by instruction works: evidence for productive failure. Review of Educational Research. doi:10.3102/00346543211019105
  4. Deslauriers, L. et al. (2019). Measuring actual learning versus feeling of learning in response to being actively engaged in the classroom. PNAS 116(39). doi:10.1073/pnas.1821936116
  5. Roediger, H. L. & Karpicke, J. D. (2006). Test-enhanced learning. Psychological Science. doi:10.1111/j.1467-9280.2006.01693.x
  6. Dunlosky, J. et al. (2013). Improving students' learning with effective learning techniques. Psychological Science in the Public Interest. doi:10.1177/1529100612453266
  7. Hopkins, R. F. et al. (2016). Spaced retrieval practice increases college students' short- and long-term retention of mathematics knowledge. Educational Psychology Review. doi:10.1007/s10648-015-9349-8
  8. Bisra, K. et al. (2018). Inducing self-explanation: a meta-analysis. Educational Psychology Review. doi:10.1007/s10648-018-9434-x