We built an LMS. Then we understood what we’d actually built.

HabileLabs has built cloud, data and AI platforms for banks, NBFCs, manufacturers and energy companies since 2014. Across all of them, one thing kept showing up in the corner of the room: training was the weakest tool anyone owned.

The pattern we kept seeing

Safety procedures learned from a PDF on a shared phone on the shop floor. A compliance LMS at an NBFC that employees had stopped opening. And every audit, the same scramble — attendance emails, scanned sign-in sheets, a spreadsheet someone updated by hand — assembled over a week to answer one question: are our people trained?

So we built Learnify. We built it as an LMS, because that is what the category is called, and for a while we competed the way the category competes: a better course builder, a nicer learner app, more features.

The week before the audit

What changed our mind was watching how customers actually used it.

Nobody called us the week a course launched. They called the week before an audit. And what they asked for was never an engagement feature. It was: who has expired? Who has not attested? Can I lock access when a certification lapses? Can I export exactly this for the auditor, today?

What we built instead

Recertification cycles anchored to each person’s own completion date, not a company-wide reset. A read-and-understood attestation for the policies where the law expects a signature, not a score. Access that re-locks when a certification lapses, and renewals that queue for approval. An audit log of every action. And an evidence pack — completions, scores, attestations, certificates, all timestamped — that exports in one click.

Then India’s Digital Personal Data Protection Act arrived, and our customers asked a new version of the same question: can you prove consent? Can you track a rights request? Can you show we trained everyone on this? We built that too — as regulatory profiles, so the same machinery serves GDPR for our European customers, and whatever comes next is a new profile rather than a new platform.

When we added AI, the first rule we wrote was about what it must never do. Our three agents — the L&D Copilot, the Compliance Watchdog, the Learner Tutor — draft courses, watch validity and answer learners’ questions. None of them approves a course, grants access or certifies a person. On a platform whose job is proof, an agent that decides would be an audit finding, not a feature.

Not a learning system. A proof system.

Somewhere in that work we understood what we had actually built. Not a learning system. A proof system. And once we said it out loud, a number of decisions became simple.

What it changed

• We collapsed our pricing into one plan with every feature in it, because gating the proof layer behind an upsell would undercut the only thing we are really selling.
• We published the prices, with a calculator, so you can work out your own number before you talk to anyone.
• We opened a free trial — twenty seats, thirty days, everything on.
• We went through our own website and removed every claim we could not point to in the product. A strange discipline for a marketing exercise to leave you with, but the right one for a company whose product is proof.

An invitation

If you are the person who has to show an auditor, a client or a regulator that your people are trained — and still current — I would genuinely like to hear how you do it today. What is the slowest part? What breaks? Write to us, or find me on LinkedIn.

And if you would rather see it than read about it, Learnify is free to start. Upload your worst policy PDF and watch what happens.