Glossary

The words we use

We use some words in a particular way, and a few of them are Sanskrit or Marathi. Here is what each one means in plain English. If a word on this site is not on this page, it is meant in its ordinary sense.

Our name

The first word we use is our own.

LOFL: The company. Read it however you like. The registered company is LOFL (OPC) Private Limited. We read the four letters as Leap of Faith Labs. Other people have read them as Language of Laughter, Letter of Love, Love of Life. L-OF-L holds whatever you bring to it.

The path a piece of work takes

Every project on this site moves along the same six steps, in order.

The spine: The six steps, start to finish. Activity → Evidence → Verification → Tokenization → Exchange → Regeneration. We do not skip a step, and we do not treat an activity as its own reward before we have said what would count as evidence.
Activity: Someone actually does something. A real human act of learning, teaching, creating, stewarding or caring. We help design it so it can be finished and observed.
Evidence: Proof it happened. Artefacts, logs, assessments or records. We help define what counts, and how it is captured and stored — with consent.
Verification: Somebody checks the proof. A human or AI-assisted review against criteria set in advance. We design the rubrics and flag the gaps.
Tokenization: The check is written down. An Impact Unit is issued to the contributor's ledger. Issuance stays conservative and earned.
Exchange: The record becomes useful. Verified impact moves into a marketplace, a recognition system or a policy layer — never conflating internal units with regulated financial instruments.
Regeneration: Value comes back to the people who made it. The loop that makes the system self-sustaining. We name how value returns to contributors and communities, and flag extractive patterns.

What gets written down

Three words for the record-keeping.

Impact Unit: One verified thing someone did. A credential and a reputation signal — not a security, and not tradable. One activity issues one primary unit; we never stack units on a single unverified activity. Four kinds: Learning (LIU), Skill (SIU), Cultural (CIU) and Green (GIU).
The three ledgers: Three books the work gets recorded in. Knowledge (resources, pathways, curricula, mentor time), Impact (verified claims, milestones, evidence packs) and Growth (capability and governance maturity). We always know which book we are writing to.
Verification levels: How hard the proof was checked — L0 to L3. L0 self-reported, and flagged as unverified. L1 checked by peers or mentors. L2 checked by a qualified third party. L3 checked through a regulated workflow. We start at the simplest level the stakes allow.

The four tests

Every deliverable has to pass these before it ships.

The Sita Tai Test: Could a semi-literate village teacher use this without humiliation? Named for a woman in a Konkan village who teaches literacy to five children at home. Could she take part, have her contribution verified and receive real recognition — without humiliation, excessive friction, or depending on systems she cannot reach or afford?
The 15-minute steward test: Can a teacher run this in under fifteen minutes a week?
The ledger-ready test: Which book does this belong in, and what would make it ready to go in?
The VIC test: Does this get us an actual verified claim, or just a nicer framework for making one someday?

The program names

Sanskrit and Marathi names, and who each one is for.

Sāṃvāda: A conversation between equals — the asker answers first. सांवाद. Five questions we have already answered in public before we ask you. Skipping a question is an answer. Answers come back as a set, side by side and whole — never averaged, never scored.
Nāgarika: For civic-tech platforms. नागरिक — citizen. If you run a platform citizens depend on, and AI is entering it faster than your judgement of it.
Sahakāra: For cooperatives and collectives. सहकार — co-operation. If your members own the institution, and you want them to own its AI decisions too.
Vivekī: For mid-sized NGOs. विवेकी — one who discerns. If funders are asking about your AI posture and you would rather have discernment than a policy PDF.
Hastakāra: For artisan clusters and women-led MSMEs. हस्तकार — maker by hand. If the cluster's craft is the asset, and AI should serve the makers, not describe them.
Svarāj: The open individual cohort. स्वराज — self-rule. Eight weeks, at full speed, with judgement intact. Deliberately not an efficiency course.
Signal: Where you tell us what did not work. No ratings anywhere; your words stay whole.

How the tools behave

Words that describe what the tools do and refuse to do.

Zero-egress: Nothing you type leaves your device. No accounts, no tracking, no backend. The page keeps what you make, on your machine.
Sets and tallies, never scores: We count things; we never rank people. Badges and traces accumulate as a set. There is no leaderboard, no streak and no grade.
You press the mark: A person confirms it, not the software. Agents propose; policy decides. No agent writes to a verified record without a human sign-off.
A trace: One honest line about what you actually did.
An absence: Something the page cannot show, named on its face. On Atlas plates, what is missing gets a number and stays visible — an English-only harvest, an un-consulted archive. We keep the honest fail rather than hiding it.