Voice waveform transforming into layered data streams
Adaptive Responsive AI for Speaking Hub

Every spoken word becomes measurable progress

ARASH is an enterprise English speaking platform for junior high school. Students speak, eleven AI engines analyse, and students, teachers, parents, researchers and administrators each get the view they need — with evidence they can play back.

  • Consent-gated AI for minors
  • Evidence-linked scoring
  • Teacher override on every score
  • Analytics for schools and districts
6
skill dimensions scored
11
specialised AI engines
<20s
feedback turnaround
5
role-specific workspaces

How it works

One recording, a full feedback loop

A student speaks once. ARASH transcribes, scores six dimensions, writes feedback for each audience, and reshapes tomorrow's practice.

  1. 01

    Student speaks

    A prompt, a task or a live conversation. Recording uploads in resumable chunks.

  2. 02

    AI engines analyse

    Transcription, pronunciation, fluency, grammar, vocabulary, content and interaction — in parallel.

  3. 03

    Feedback is composed

    Audience-specific wording for the student, the teacher and the parent, from one score set.

  4. 04

    The path updates

    The learner model shifts, tomorrow's plan changes, and every dashboard refreshes in realtime.

Platform

Built for how speaking is actually taught

Practice, assessment, evidence and analytics in one place — not four disconnected tools.

Speaking practice that adapts

Daily plans rebuild themselves after every attempt, targeting the dimension holding a learner back instead of a fixed syllabus.

Live AI conversation partner

Open-ended dialogue at the learner's level, with scaffolding that fades as fluency grows and a transcript teachers can review.

Six-dimension assessment

Pronunciation, fluency, grammar, vocabulary, content and interaction — each with a band, a trend and the exact audio moment behind it.

Evidence you can play back

Every score links to a timestamped clip, so feedback is a conversation about something real rather than an opaque number.

Class and cohort analytics

Skill heatmaps, participation, and early-warning flags for teachers; enrolment, usage and outcome trends for administrators.

Privacy built for minors

Short voice retention, parental consent scopes, pseudonymised research exports and a full audit trail on every access.

Assessment

Six dimensions, one composite band

Each score links back to the exact moment in the recording that produced it, and every teacher override is tracked as a revision.

Pronunciation

78

Grammar

71

Vocabulary

83

Fluency

66

Comprehension

88

Pragmatic

74

For everyone involved

Five audiences, one source of truth

The same score set becomes a practice plan, a class heatmap, a parent digest, a research row and an admin metric.

Students

Know exactly what to practise next

  • A daily plan under 15 minutes
  • Instant, kind, specific feedback
  • Playback of their own best takes

Teachers

Intervene before a student falls behind

  • Class health at a glance
  • Assignment delivery and review
  • Override any AI score with a reason

Parents

Follow progress without jargon

  • Weekly plain-language digest
  • Consent and privacy controls
  • Direct line to the teacher

Researchers

Study speaking development at scale

  • Pseudonymised, append-only datasets
  • Rubric and model version stamps
  • Reproducible export manifests

Administrators

Run the programme with confidence

  • Tenancy, roles and retention
  • Usage and cost visibility
  • Audit trail for every action

Ready to hear your students speak?

Start a free classroom pilot, or talk to us about a school-wide rollout.