Most team-training tools end at a completion dashboard: who clicked through, who didn’t. That tells you a course was finished — not that anyone will still know the material six weeks from now, which is the only week that matters. Quizlar for Teams is built around the second question.
Learners study hands-free by voice or on screen, against the same FSRS spaced-repetition engine that powers individual Quizlar accounts. The organization gets shared decks, roles, SSO, central billing — and an admin view that reports outcomes rather than attendance.
Org analytics report outcomes: completion, mastery, last reviewed, and cards remaining, per learner and per deck. Not answers, not transcripts. A manager can see that a policy deck is 40% mastered across the team and which sections lag — without reading what any individual typed or said into a practice question.
That distinction matters more than it sounds. Training analytics that surface raw answers turn every practice attempt into a performance record, and learners respond by not practising honestly. Reporting mastery instead keeps the retrieval practice safe to get wrong, which is the entire mechanism by which it works.
— Sangura, founder of Quizlar. I build it solo, so I’m allergic to tools that quietly turn every teammate into an unpaid card-maker.
Roll this up to a team and the friction compounds. Most training makes each person build their own recall — take notes, remember to review, self-assess — and predictably almost nobody does. The completion dashboard looks fine; six weeks later the knowledge is gone. It’s busywork you’d never assign on purpose, baked into the tool.
I build Quizlar solo — I started it in December 2025, once agentic coding tools got good enough to build the learning tool I’d always wanted. My rule for it is simple: learning is a mental gym, so the app can rack the weights, but you still have to lift. Any tool that quietly hands you back its own job — making the cards, grading yourself, staring at a screen — is a tool you’ll quit. That’s the one thing Quizlar won’t do.
The reason internal training decks never get built is that somebody has to sit down and write the questions. That is the work Quizlar takes off an author's desk. Paste a policy document, a runbook, a product FAQ, the notes under a slide deck, an article URL, a PDF, or the link to a recorded training video, and the deck comes back as questions and reference answers in about thirty seconds. The author's job shifts from writing cards to reviewing them — deleting the ones that test trivia, tightening the reference answer where the wording matters legally, and publishing the rest to the organization.
A few things separate an org deck that works from one that quietly rots:
Authors edit in the same workspace learners study in, so there is no separate authoring tool to license or learn.
A learner opens a deck on a phone, in a car, or while walking, and Quizlar reads the question aloud. They answer in their own words. Speech recognition transcribes it, and an AI grader compares the answer against the card's reference answer — not as a string match, but for whether the substance is there. Say the right thing in the wrong order and it counts. Miss the part that matters and it does not.
Three behaviours follow from that, and they are the reason spoken practice beats multiple choice for training material:
Voice is optional throughout. The same deck works typed, on screen, for anyone who prefers it or is somewhere they cannot talk. Quizzing works in 60+ languages, which matters for distributed teams whose working language is not everyone's first one.
Be clear-eyed about the grader: it is a language model, and on a badly worded reference answer it can mark a correct answer wrong. That is why authors review reference answers, and why analytics report mastery trends rather than treating any single graded card as a record about a person.
Teams is seat-based with a three-seat minimum, billed monthly or annually to one invoice that an admin manages. Every seat brings 150 voice minutes per month, pooled across the organization — a heavy week for one learner draws on the same pool as a quiet one for someone else, so you are buying the team's usage rather than rationing each person's. Past the pool, voice keeps working and is metered at $0.10 per minute. Screen-based study is not metered at all. Current per-seat pricing lives on the pricing page.
The practical planning number is the pool, not the seat count: a fifteen-minute review habit twice a week runs roughly 120 minutes a month per active learner, so a team where everyone studies by voice will use most of its allotment, and one where half the team studies on screen will not come close.
Worth knowing before a pilot, because a tool's edges cost more to discover later:
New-hire onboarding, product and feature knowledge for sales and support, compliance and safety training, certification prep, and keeping a distributed team aligned on fast-changing information. Anywhere "they were told once" isn't good enough.
A single training session is forgotten fast — the classic forgetting curve. Spaced repetition resurfaces each item at expanding intervals, converting short-term exposure into durable memory with a fraction of the total study time. For an organization, that's the difference between a compliance checkbox and a workforce that actually knows the material.
If you want to see it against your own material rather than a demo deck, the fastest test is a single deck built from the one document your team is most often wrong about, given to three or four people for a fortnight. Two weeks of real reviews tells you more than any feature list — including, sometimes, that the material itself was the problem.
Quizlar for Teams is the organization tier: shared decks, admin/author/learner roles, per-deck cohort analytics, SSO, and central billing, all powered by the FSRS spaced-repetition engine so team training is actually retained.
Three: admin (manage members, billing, and settings), author (create and edit shared org decks), and learner (study the org's shared decks and track progress). There is no separate owner role; billing is handled by an admin.
Yes — as outcomes. Org analytics show completion, mastery, last reviewed, and cards remaining, per deck and per member, so you can see what's landed and where people are falling behind. They do not show what a learner answered: no answer text, no voice transcripts.
Yes. Every organization's data is scoped and isolated at the database layer using row-level security, so one org can never access another org's decks, members, or analytics.
Yes. Connect your identity provider over OIDC and members sign in with their existing work accounts. OIDC works today; SAML we’d build for a pilot. See the SSO guide for setup.
Through central billing managed by an org admin — one account for the whole organization instead of individual subscriptions.
Every purchased seat brings 150 voice minutes per month and the minutes are pooled across the organization, so one learner's heavy week draws on the same pool as another's light one. Beyond the pool voice keeps working and is metered at $0.10 per wall-clock minute. Studying on screen is not metered.
No. Quizlar has no assignments, due dates, SCORM packages, or certificates — it is the retention layer that runs after the course. Teams typically keep the LMS for delivery and compliance records and use Quizlar so the material is still known months later.
An author pastes a policy document, runbook, article URL, PDF, or training-video link and Quizlar generates questions with reference answers in about 30 seconds. The author reviews and edits before publishing to the organization, so the work is editing rather than authoring from scratch.
SSO runs over OIDC today, which covers Google Workspace, Microsoft Entra ID, Okta, and most identity providers. SAML is not shipped; we would build it for a pilot that requires it.