Speech, measured the same way twice.

For research laboratories

An instrument is only worth what its numbers can be trusted to say. Ours have not been checked — and that is what a founding partnership is for.

Speech is measured three ways today. A clinician scores the session live, which is reliable but needs an examiner and a schedule. A recording is scored by hand, which is what most laboratories do — slow, and the score depends on who did it. Or software scores a tightly scripted task and returns a number without the record behind it.

SonaXR runs the session itself: the same voice, the same prompts, the same pacing, for every participant. Scoring is deterministic — compared against a fixed reference, or against a key a linguist set in advance. Nothing is scored by hand. And the whole record comes out: every item, every response, every timing, in open formats. A run six months later is comparable to this one.

Six fields already depend on measuring speech

None of these is a validated protocol. Each is a study a laboratory could run today, with the content authored to its own constructs.

The scientific basis for each design decision is set out in the Foundations brief — one peer-reviewed construct per implementation choice, including where the evidence is mixed.

Second-language acquisition and sustainment

Proficiency is perishable between formal evaluation cycles, and the measurement in between is usually nothing at all.

This is the same instrument, pointed at a language learner rather than a research participant. The session is administered identically every time, without an examiner and without a schedule, which means a reading taken in March is comparable to a reading taken in September. That comparability is the product. It is measurement between formal evaluations — not a substitute for one.

Commander's Package

$14,500 annually

Covers

25 personnel, three measurement rounds per year

Delivery

Twenty minutes, seated, on a commercial headset a unit can already own. No workstation, no wires, no sound booth, no examiner.

What comes back

Every item, every response, every timing, in open formats — the same record the research program produces.

Placement

On a ladder referenced to the Interagency Language Roundtable descriptors. The correspondence to an official Interagency Language Roundtable rating is unvalidated; establishing it is what the research partnerships are for.

$14,500 sits below the federal micro-purchase threshold of $15,000. Pricing is a hypothesis — nobody has paid it yet.

Why an instrument that has not been validated is still worth deploying between evaluations: see where the existing approaches stop, and what we are not claiming.

What the instrument could be pointed at next, once a partner defines it

The six fields above exist because researchers described them. A seventh happens the same way: a laboratory defines the construct, and the content is authored to it.

Every approach here exists but one.

Spoken proficiency can already be measured by machine, and has been for decades. Knowing precisely where each established approach stops is the entire basis of what we built.

How spoken-proficiency testing divides by who delivers the session and who scores it. SonaXR occupies the only combination with no existing approach.

	A human rates it	A machine scores it
A person delivers it	The certified interviewThe gold standard. Twenty to thirty minutes, one trained tester, functional tasks, a rating on a recognized scale.	No existing approach—
A computer delivers it	Avatar-delivered interviewPrompts on demand, in many languages. Keeps the functional tasks — and pays for it, because every recorded session still goes to certified human raters.	Automated scripted-task scoringNo human rater, results in minutes, in production since the 1990s. Works by keeping tasks tightly controlled — which is also why they measure something other than functional language use.
A computer delivers it inside a situation	No existing approach—	SonaXRFunctional speech in a situated scenario, scored without a rater, on a record that carries no identifier mapping back to the speaker.

The gap, in one sentence. Automated scoring is bought by giving up functional tasks. Functional tasks are kept by paying humans to rate them.

Why a headset, when existing tests need only a phone

The honest answer is not immersion for its own sake. Repeating a sentence into a handset measures something real, but it is not the same construct as handling a situation in the language. Functional tasks need a situation to be functional in, and virtual reality is how that situation is constructed identically for every participant, every session, at every site. The headset is the fixture that holds the conditions constant. It is a means, not the product.

What makes a task functional rather than scripted

A scripted task fixes the correct answer in advance — repeat this sentence, read this passage. A functional task fixes the goal and leaves the wording open: describe what happened, narrate this sequence, handle this situation. Those are the tasks the recognized proficiency scales are written around, and they are hard to score automatically for exactly the reason they are worth scoring. SonaXR narrows that space without scripting it — the situation constrains what a reasonable answer looks like, and scoring runs against a frozen reference and answer sets a linguist approved in advance, rather than against free transcription of accented speech.

The instrument, in parts

Status: pre-seed prototype. What follows is running code demonstrated on Meta Quest 3 hardware, not a shipped product.

Delivery

Virtual reality on a commercial standalone headset. Meta Quest 3 is the primary and only validated target; built on Unity XR for portability, but current builds rely on Meta-specific hand tracking and other headsets are untested.

Speech

Phoneme-level pronunciation scoring through a commercial speech service. Adding a language is content work rather than a research program.

Scoring

Deterministic — compared against a fixed reference, or against a key a linguist set in advance. Nothing scored by hand, nothing interpreted.

Coaching

A test that helps you is not a measurement. Coaching was removed at the source-code level, and the build verifies it is absent every time.

Record

Every item, every response, every timing, in open formats — structured, audit-logged, and exportable into an organization's own analysis tools.

Language coverage

The pipeline is language-agnostic by design, and a demonstrated English scenario runs on the same architecture that scores European Portuguese. English is not a separate product; it is the instrument pointed at a different target language.

Two people

One engineer, one operator. The team that built SonaXR is the team that operates it.