Echo voices and Baobab national models — Africa-first language intelligence.

Explore Baobab
Skip to content
Open consoleLog in

Lugemi status

Accents engine

Accent identity engine catalogue and regional profiles.

All status pagesRelated viewAll status pagesMachine JSON
product
Lugemi Accent Intelligence
note
Cue-based spoken accent detection/classification with confidence and analytics. Dialect detection is Language Cloud. Regional acoustic models are deferred.

capabilities

capabilities: 7

  • Accent Detection

    id
    accent-detection
    name
    Accent Detection
    status
    shipped
    api
    POST /v1/accents/detect
    notes
    Text and/or audio→STT cue scoring. Not acoustic phonetics ID.

  • Accent Classification

    id
    accent-classification
    name
    Accent Classification
    status
    shipped
    api
    POST /v1/accents/classify
    notes
    Ranked candidates + confidence band from the same cue engine.

  • Accent Confidence

    id
    accent-confidence
    name
    Accent Confidence
    status
    shipped
    api
    POST /v1/accents/detect
    notes
    Per-result confidence + candidate scores.

  • Dialect Detection

    id
    dialect-detection
    name
    Dialect Detection
    status
    shipped
    api
    POST /v1/dialects/detect
    notes
    Language Cloud — linked here, not reimplemented under Speech Cloud.

  • Regional Accent Models

    id
    regional-models
    name
    Regional Accent Models
    status
    deferred
    api
    null
    notes
    Acoustic / phonetics regional models — buy path / later depth. Not claimed.

  • Accent Analytics

    id
    accent-analytics
    name
    Accent Analytics
    status
    shipped
    api
    GET /v1/accents/analytics
    notes
    Org audit-derived detect/classify counts (30-day window).

  • Monitoring

    id
    monitoring
    name
    Monitoring
    status
    shipped
    api
    shared observability
    notes
    Request IDs + accent.detect / accent.classify audit actions.

engines

engines: 1

  • Lexical cue scoring (+ optional LLM assist)

    id
    cue_scoring_v1
    name
    Lexical cue scoring (+ optional LLM assist)
    role
    primary

    modes

    modes: 2
    • detect
    • classify

regionalProfiles

count
seeded
note
Full language/locale spoken accent registry generated from languages + locale/country packs — cue scoring, not acoustic models.
console
/accents

links

architecture

rest
true
graphql
true
sdk
@lugemi/sdk
cli
@lugemi/cli
docker
true
terraform
true
kubernetes
true
primaryRegion
af-south-1
deployment
Fly default; optional EKS af-south-1 (shared platform)