You cannot hear forty students individually
A ninety-minute class split forty ways is two minutes each. Most students finish a course never having been corrected on the sound actually holding them back.
For academies, institutes & coaching centres
Every student gets a pronunciation lab. Every teacher gets the evidence.
Nadiv scores your learners' speech sound by sound — all forty-four English phonemes, on every attempt — decides what each of them should practise next, and rolls the whole cohort up into a console that shows you who has stalled and which sounds to spend Monday's class on.
- 44 phonemes scored individually
- One join code per class
- English first, more languages following
- Priced per institute
The problem
Pronunciation is the one thing a classroom cannot scale
Not because teachers are not good enough. Because the work is per-student, per-sound, and endlessly repetitive — three properties a room full of people does not have.
Drilling eats the class they paid for
Repetition is where pronunciation is won, and the least valuable possible use of a trained teacher in a room.
Progress is anecdotal
“She has improved a lot” is the report you can give a parent today. No baseline, no number, nothing tying the gain to your course.
Adaptive engine
It decides what every learner practises next
This is the part a teacher cannot do forty times a week. Nadiv holds a mastery profile for every sound each learner has attempted, and opens every session on the one costing them the most accuracy today — converting passive repetition into targeted practice.
Next for this learnerSample data
/θ/as in “think”
Mastery41
/θ/ is the lowest-scoring sound in this learner's profile, at 41. Focused reps here move their overall score more than anything else available.
Ranked from every drill targeting this learner's weakest sounds
- Voiceless TH — word-initialBest fit — targets the weakest sound, not yet attempted/θ//s/0.71
- TH vs S minimal pairsStrong fit, held back for the next session/θ//s/0.64
- Fricative ladder — mediumCovers the target but dilutes across three sounds/θ//ð//v/0.58
- Voiceless TH — connected speechAlready completed — sinks in the ranking/θ/0.52
No syllabus to re-sequence by hand, and no two learners on the same path unless their profiles say they should be. How the engine works.
Getting a class running
Three steps, and no IT project
The whole rollout is one code per class. Nothing to install on your network, nothing to sync, nothing to maintain.
- 01
Create a classroom, get a code
Name the batch in the teacher console. Nadiv issues a six-character join code drawn from an alphabet with no 0, 1, I, O, L or U in it — so it survives being read aloud to a room or written on a whiteboard.
- 02
Students redeem it in the app
One code for the whole cohort. No spreadsheet import, no accounts to provision, no IT ticket. A student installs Nadiv on the phone they already carry and types the code once.
- 03
Figures arrive as they practise
Every attempt is scored at the sound level and aggregated into the roster. A student who has joined but not yet drilled shows dashes and zeros — that is correct rather than broken, and it is the single most useful row on the page.
The teacher console
What you see on Monday morning
One row per student, ordered so the students who need chasing are at the top of the page rather than buried in the middle of it.
ClassroomsIELTS Speaking — Batch A
Join code4KF 2MW
| Student | Drills | Attempts | Avg score | Improvement | Last active |
|---|---|---|---|---|---|
| Kenji T. | 0 | 0 | — | — | Never |
| j.mensah@example.com | 1 | 3 | 31 | — | 3 weeks ago |
| Priya N. | 4 | 12 | 49 | -1 | last week |
| Tarek B. | 9 | 31 | 66 | +2 | 4 days ago |
| Sofia D. | 15 | 58 | 74 | +6 | 2 days ago |
| Rohan M. | 18 | 74 | 87 | +9 | yesterday |
| Aisha K. | 24 | 96 | 91 | +12 | 2 hours ago |
Ordered by who has fallen behind
Least-recently-active first, because “who has stopped practising” is the question you open this to answer — not “who is top of the class”.
Improvement is measured from the day they joined you
Each learner's score is captured the moment they redeem your code. The delta on the roster is the progress your programme is responsible for, never a lifetime total that predates it.
Open one learner and you get their sounds
Score trend, a thirty-day practice pattern, their weakest phonemes with example words, and the sounds that have moved most since joining.
IELTS Speaking — Batch ASofia D.
Practised score74
Score trend
Last 20 scored attempts, oldest first
Practice, last 30 days
One block per day, shaded by attempts
Weakest sounds
Where tomorrow's lesson should go
- /θ/think41
- /ð/that46
- /v/very52
- /ɹ/red58
- /æ/cat61
Most improved
Against her score when she joined
- /ɹ/red+34
- /v/very+22
- /æ/cat+14
A student who has joined your class but not yet drilled shows dashes and zeros across the row. That is the system being honest rather than broken — and it is the row worth acting on first.
Advantages
Why this rather than what you do now
Measured against the three real alternatives: a teacher's ear, a generic speaking app, and a language lab in a room.
Phoneme-level, not sentence-level
Most speech tools return one score for an utterance — “roughly 72% intelligible” — which gives a learner nothing to act on. Nadiv localises the failure to the sound: /θ/ at 41, /ɹ/ at 78. That is the only resolution at which pronunciation is actually fixable.
The engine decides what each learner does next
Not a fixed syllabus every student walks through in the same order. Nadiv profiles each learner's mastery sound by sound and sequences their next drill against it, so practice time goes to the sound costing them the most accuracy today.
Learners are told how, not just how well
A score alone does not teach anyone to produce /θ/. Articulatory coaching gives tongue placement, lip rounding and airflow per sound, with minimal-pair loops for the contrasts only audible in comparison.
Homework becomes observable, and the roster stays yours
Drills, attempts, last active and a thirty-day pattern per learner — the practice you set outside class stops being something you take on trust. Learners join only by redeeming your code, and you see only the classes you created.
Commercials
Priced per institute, after we have talked
There is no price table on this page and that is deliberate. A forty-student coaching centre running a twelve-week IELTS cycle with two teachers, and a nine-hundred-student university department running a full academic year, are not the same product. Charging them from the same list means one of them is overpaying.
So we quote after a conversation. It is usually a short one, and you will get a number rather than a range.
What changes the number
- How many students are in the cohort
- How many teachers need a console
- How long your term or course cycle runs
- How much rollout and teacher training you want from us
- Whether you need something built that does not exist yet
Included in every arrangement
What you get, whatever the number is
- The Nadiv app for every student in the cohort, on iOS and Android
- A teacher console for each of your teachers, with unlimited classrooms
- The full forty-four-phoneme lab: mastery grid, drill loops, minimal pairs, mouth-position coaching
- Per-student insights: score trend, practice activity, weakest sounds, most improved
- Onboarding for your teaching staff and a rollout plan for the first cohort
- A named contact who answers email
Book a demo
Tell us about your institute
A few lines is plenty. We will come back with a walkthrough of the console using your own class as the example, and a number that fits your cohort.