The five questions (Sāṃvāda)
Five questions about your work and AI. Our founder’s answers are already there. Answer in any language, at any length. Skipping a question is also an answer.
Programs › AI for ALL
AI for ALL · a Leap of Faith Labs initiativeAI for ALL helps you think clearly about AI in your own work: where it helps, where it does harm, and what must stay human. It is for one person, a teacher, a parent, or a whole organisation. Saying no to AI is a valid answer here.
Each one takes you to the right part of this page.
Start with the free Svarāj workbook. One-to-one Svarāj opens later.
SvarājThe Staffroom Loop workbook: ten hours on your own lessons, one hour at a time. Free.
Free workbooksHow we work with children and young people, and how we keep them safe.
Children and young peopleSchools start the same way as every organisation: five questions, then a conversation.
How to beginBuild real judgement in your team, not just a policy document.
VivekīFor cooperatives, producer groups and federations.
SahakāraFor platforms and projects that serve citizens.
NāgarikaFor artisan clusters and networks of women-led small businesses.
HastakāraFund a program for a craft cluster, or a pilot through to its first checked result.
For fundersNo sign-up, and nothing tracks you. What you type stays on your device.
Five questions about your work and AI. Our founder’s answers are already there. Answer in any language, at any length. Skipping a question is also an answer.
Ten working days on one piece of your own work: what AI does inside it, and what it never does. Write on screen or print it on A4.
Ten hours on your own lessons, one hour at a sitting. Every hour leaves you something you can hold. Nothing is scored.
Nine plain cards to say what AI did in a piece of work, what stayed yours, and what you checked. Fill one in and show it.
Everything here is shared under CC-BY-SA 4.0 — a Creative Commons licence. In plain words: you can use it, change it and share it, as long as you name where it came from and share your version just as freely. More free workbooks are on the Pathika shelf.
Every AI for ALL program — for one person or a whole organisation — follows the same four steps.
A conversation between equals. Whoever asks, answers first. We start by listening to how your work really runs — not the official version.
Together we map your work: which tasks AI could touch, where it would help, where it would harm, and what must stay human. We look at the evidence, and at the strongest case against.
We try the smallest real use that can show whether it works, and check it for harm. If the honest answer is “don’t use AI here”, that is a full result, not a failure.
The work is written up so others can learn from it. You decide what happens next — and you judge us.
Every price is public. Prices are in rupees, tax included, and every engagement runs on a signed letter.
For you if you want to use AI at full speed and keep your judgement. It is deliberately not a course about going faster.
Join on WhatsApp — message ready Or start with the five questions
For you if citizens depend on your platform, and AI is arriving faster than your judgement of it.
₹1.5–4.5 lakh advisory · ₹4–16 lakh embedded
For the first pilots, we write up the work in public in place of a fee. The first pilot is underway.
For you if your members own the institution, and you want them to own its AI decisions too.
₹1.5–6.5 lakh advisory · ₹4–16 lakh embedded · ₹3–12 lakh train-the-trainer
A small fee always — never free, because what is free is never truly owned. Travel is charged at what it costs.
For you if funders are asking about your AI position, and you would rather have real judgement than a policy PDF.
₹75,000–2 lakh Discernment Lab · ₹5–12 lakh facilitator training (8 weeks) · ₹7–17 lakh both, with six months of follow-up
Padtalani, our AI impact check in line with India’s DPDP Act (the Digital Personal Data Protection Act, 2023): ₹75,000–3 lakh on its own, by the size of your organisation.
For you if the craft is the asset, and AI should serve the makers rather than describe them. For artisan clusters and networks of women-led MSMEs (micro, small and medium enterprises).
₹15–30 lakh full program, paid by a funder
This covers field costs too, including paid time for the artisans who take part. We can invoice in US dollars for international funders.
Not sure which one fits? Tell us the size of your organisation and what you are deciding, and we will send you what it costs. Send an inquiry →
Hastakāra is paid for by a funder, so the makers do not carry the cost. You can also fund any program as a pilot, through to its first checked result — a real outcome that someone else has verified, not a deck describing one. The documentation is open, and the upside is shared with the communities involved.
Pick what fits, and we write the message for you. You read it, then send it on WhatsApp or by email yourself. This page stores nothing and sends nothing.
For schools, NGOs, cooperatives, platforms and clusters. Either side can stop at any step.
Schools start the same way. Anything that involves children follows our child safeguarding policy.
Many children already use AI, often with no adult beside them. We work with children through the adults who care for them — a parent, a guardian or a teacher.
These hold in every program, for every person.
No. You need your own work and an honest look at it. We start from what you already do.
Yes. Deciding where AI does not belong is half the work. Here, refusing is counted as a skill.
No. It is about keeping your judgement while you use AI — or while you decide not to.
No. We do not sell a product, and we do not lock you in. We build on open tools, and what we make together is yours to keep and share.
The workbooks, cards and tools are free for everyone. Guided programs are paid, because what is free is rarely owned. On your own, one-to-one Svarāj is ₹10,000 for ten working days. For some first pilots, we write up the work in public in place of a fee.
Answers go back to your group as a set — side by side and whole, never averaged and never scored. You choose how you are named: your full name, your role only (for example, “a science teacher”), within the group only, or no name at all.
Children take part through a parent, guardian or teacher, and only with written consent. See children and young people.
Leap of Faith Labs (LOFL), a small studio in Pune, Maharashtra. Programs are led by our founder, Mrunmai Patil.
Use Signal. It takes about three minutes. Telling us what went wrong is the only required part — and your words stay exactly as you wrote them.
A short note is plenty. Write in any language.
WhatsApp is run by Meta, and your message travels on their service — see the privacy notice. For a child’s safety, write to childcare@lofl.world.