Listing Thumbnail

    Ekinox French & Arabic Pronunciation Assessment for SageMaker

     Info
    Sold by: Ekinox 
    Deployed on AWS
    Free Trial
    Deploy phoneme-level French and Arabic pronunciation scoring on SageMaker. All audio stays in your AWS environment - no external API calls, no third-party data transfers.

    Overview

    OVERVIEW

    The Ekinox French and Arabic Pronunciation Assessment Engine is a full-stack solution for autonomous language learning. Beyond a standard ML model, it combines audio-to-phoneme conversion with a business logic layer that handles pronunciation scoring, fault classification, utterance acceptance, and JSON diagnostic composition. Delivered as an Amazon SageMaker Model Package, it runs entirely within your AWS account. No audio data leaves your environment and no external API calls are made.

    A free trial is available so you can evaluate the engine before committing.

    KEY FEATURES

    • Full-Stack Assessment Engine: Incorporates a dedicated business logic layer for pronunciation scoring, fault classification, and utterance acceptance - all within a single deployable SageMaker Model Package.

    • Phoneme-Level Feedback: Delivers granular scoring for French and Arabic, enabling learners to identify and correct specific pronunciation faults at the phoneme level.

    • Completely Self-Contained: All audio processing, scoring, and diagnostics occur within your AWS environment. No outbound data transmissions. No external API dependencies.

    • Optimized for Real-Time Use: Engineered for low-latency inference so end-users receive immediate feedback during learning sessions.

    • Comprehensive JSON Diagnostics: Each assessment returns structured results including overall scores, accepted utterances, and categorized fault details.

    HOW IT WORKS

    The engine is delivered as an Amazon SageMaker Model Package. Once deployed to a SageMaker endpoint, you send audio input along with the target utterance and receive a detailed JSON response. The ML model converts audio to phoneme sequences. The business logic layer then scores accuracy, classifies faults, determines utterance acceptance, and composes the full diagnostic output.

    The data flow stays entirely within your AWS account: audio input flows to the ML model, then through the business logic layer, and returns as structured JSON output - all within your SageMaker environment.

    DATA SOVEREIGNTY AND USER PRIVACY

    Unlike cloud-dependent APIs that route audio to third-party servers, the Ekinox engine processes everything within your own SageMaker environment. No audio recordings leave your AWS account during inference. No external API calls are made. You maintain complete control over your infrastructure, data lifecycle, and access policies. Data sovereignty and user privacy are preserved by architecture, not by policy alone.

    IDEAL USE CASES

    • EdTech Platforms: Integrate automated pronunciation assessment into language learning applications. The engine provides diagnostic scores and fault classifications as a tool for learner improvement, not for grading, determining access to education, or making binding decisions about learning outcomes. EU buyers should be aware that AI systems evaluating learning outcomes may fall under the high-risk category in Annex III of the EU AI Act (Regulation (EU) 2024/1689). If your deployment falls within scope, you are responsible for conformity assessment, human oversight, and end-user transparency obligations.

    • Corporate Training: Support French and Arabic language development in secure internal environments. The engine delivers pronunciation diagnostics as assistive feedback. It is not designed for employment decisions, performance evaluation, or hiring and termination outcomes. EU buyers should assess whether their deployment triggers high-risk obligations under Annex III of the EU AI Act.

    • Call Centers: Assess agent pronunciation to support coaching workflows. JSON output is designed to assist human reviewers, not to replace human judgment in decisions about agent performance or employment status. EU buyers should evaluate whether use in an employment context triggers Annex III obligations.

    Note on EU AI Act: The engine provides pronunciation scoring and diagnostic data. Whether a deployment constitutes high-risk AI under the EU AI Act depends on how buyers integrate and use it, not on the engine itself. Buyers in the EU are solely responsible for their legal assessment, conformity procedures, human oversight, and end-user transparency.

    GET STARTED

    A free trial is available on AWS Marketplace. Deploy the SageMaker Model Package in your AWS environment and begin evaluating pronunciation assessment capabilities with your own audio data and use cases. Subscribe through this listing to start your trial today.

    Highlights

    • ZERO DATA EXPOSURE BY DESIGN: Ekinox processes all speech data locally within your AWS environment on Amazon SageMaker. No audio bytes leave your infrastructure - no external API calls, no third-party data transfers. Unlike cloud-based pronunciation APIs that route audio to external servers, Ekinox keeps every inference request inside your own VPC, giving you full control over data sovereignty and user privacy from day one.
    • PHONEME-LEVEL FRENCH AND ARABIC SCORING IN ONE PACKAGE: A single deployable SageMaker model package delivers granular pronunciation assessment for both French and Arabic. The engine returns overall scores, accepted utterances, and categorized fault details in a structured JSON response - providing learners and instructors actionable, phoneme-level feedback rather than a simple pass/fail grade. Both languages are supported from the same endpoint without switching models or redeploying.
    • DEPLOY TO SAGEMAKER AND START SCORING: Deploy the model package to a SageMaker endpoint, send audio input with the target utterance, and receive a detailed JSON diagnostic response. The architecture uses a single SageMaker endpoint, keeping infrastructure simple and scalable. A free trial is available so you can evaluate scoring quality in your own environment before committing - deploy the package, test with your own audio inputs, and assess the output firsthand.

    Details

    Sold by

    Delivery method

    Latest version

    Deployed on AWS
    New

    Introducing multi-product solutions

    You can now purchase comprehensive solutions tailored to use cases and industries.

    Multi-product solutions

    Features and programs

    Financing for AWS Marketplace purchases

    AWS Marketplace now accepts line of credit payments through the PNC Vendor Finance program. This program is available to select AWS customers in the US, excluding NV, NC, ND, TN, & VT.
    Financing for AWS Marketplace purchases

    Pricing

    Free trial

    Try this product free for 7 days according to the free trial terms set by the vendor.

    Ekinox French & Arabic Pronunciation Assessment for SageMaker

     Info
    Pricing is based on actual usage, with charges varying according to how much you consume. Subscriptions have no end date and may be canceled any time.
    Additional AWS infrastructure costs may apply. Use the AWS Pricing Calculator  to estimate your infrastructure costs.

    Usage costs (4)

     Info
    Dimension
    Description
    Cost
    ml.c6i.large Inference (Batch)
    Recommended
    Model inference on the ml.c6i.large instance type, batch mode
    $100.00/host/hour
    inference.count.m.i.c Inference Pricing
    inference.count.m.i.c Inference Pricing
    $0.001/request
    ml.m5.large Inference (Batch)
    Model inference on the ml.m5.large instance type, batch mode
    $100.00/host/hour
    ml.c7i.large Inference (Batch)
    Model inference on the ml.c7i.large instance type, batch mode
    $100.00/host/hour

    Vendor refund policy

    All sales are final and we do not offer automatic refunds for this AWS Marketplace product.

    However, we at Ekinox are committed to your success. If you encounter exceptional circumstances or technical issues, we are open to discussing a resolution.

    Please contact our support team at contact@ekinox.io  with your AWS account ID and subscription details. We will review requests on a case-by-case basis.

    How can we make this page better?

    Tell us how we can improve this page, or report an issue with this product.
    Tell us how we can improve this page, or report an issue with this product.

    Legal

    Vendor terms and conditions

    Upon subscribing to this product, you must acknowledge and agree to the terms and conditions outlined in the vendor's End User License Agreement (EULA) .

    Content disclaimer

    Vendors are responsible for their product descriptions and other product content. AWS does not warrant that vendors' product descriptions or other product content are accurate, complete, reliable, current, or error-free.

    Usage information

     Info

    Delivery details

    Amazon SageMaker model

    An Amazon SageMaker model package is a pre-trained machine learning model ready to use without additional training. Use the model package to create a model on Amazon SageMaker for real-time inference or batch processing. Amazon SageMaker is a fully managed platform for building, training, and deploying machine learning models at scale.

    Deploy the model on Amazon SageMaker AI using the following options:
    Deploy the model as an API endpoint for your applications. When you send data to the endpoint, SageMaker processes it and returns results by API response. The endpoint runs continuously until you delete it. You're billed for software and SageMaker infrastructure costs while the endpoint runs. AWS Marketplace models don't support Amazon SageMaker Asynchronous Inference. For more information, see Deploy models for real-time inference  .
    Deploy the model to process batches of data stored in Amazon Simple Storage Service (Amazon S3). SageMaker runs the job, processes your data, and returns results to Amazon S3. When complete, SageMaker stops the model. You're billed for software and SageMaker infrastructure costs only during the batch job. Duration depends on your model, instance type, and dataset size. AWS Marketplace models don't support Amazon SageMaker Asynchronous Inference. For more information, see Batch transform for inference with Amazon SageMaker AI  .
    Version release notes

    Public release of the French and Arabic pronunciation assessment solution on AWS Marketplace.

    Additional details

    Inputs

    Summary

    Request format. The endpoint takes multipart/form-data — not JSON or CSV — because it carries a binary audio file alongside two text fields. Pass the complete multipart/form-data; boundary=<your boundary> string as the ContentType to InvokeEndpoint; a bare multipart/form-data without the boundary is rejected. All three parts are required and may appear in any order.

    • language (text, required) — Language of both the recording and expectedText. Accepted values: fr, fr-fr, fr_fr for French; ar, ar-sa, ar_sa for Arabic. Case and surrounding whitespace are ignored. Any other value returns HTTP 400. The response echoes the canonical short form, so fr-fr comes back as fr. Example: fr-fr

    • expectedText (text, required) — The words the speaker was supposed to say, UTF-8, in the script of the selected language. Must not be blank. Built for a single utterance — a word, a phrase, a sentence — rather than a paragraph. Example: les enfants ont mangé une petite tarte aux pommes.

    • audio (file, required) — The recording to assess: a RIFF/WAVE (.wav) file containing 16-bit PCM audio, non-empty. Any sample rate and any channel count are accepted — the model resamples to 16 kHz and downmixes to mono internally, so there is no benefit to converting first. Encoding is the one strict requirement: MP3, AAC, Opus, FLAC, and 8-bit, 24-bit or 32-bit-float WAV are rejected with HTTP 400. Convert with ffmpeg -i input.m4a -c:a pcm_s16le output.wav. Practical length is bounded by SageMaker's own InvokeEndpoint limits — 6 MB per request and a 60-second response deadline — roughly 3 minutes of 16 kHz mono audio, or about 35 seconds at 44.1 kHz stereo. Example: a 4.5-second, 44.1 kHz, 16-bit stereo WAV.

    Input MIME type
    multipart/form-data
    https://github.com/EkinoxIO/ekinox-marketplace/blob/main/pronunciation-sagemaker/data/sample_request.md
    Same as realtime.

    Support

    Vendor support

    Support Contact

    Email: contact@ekinox.io 

    Support Approach

    Ekinox provides collaborative, engineering-led support. Our team works directly with customers to understand their unique integration and operational needs rather than applying a one-size-fits-all approach.

    What We Cover

    Our engineering team is available to assist with:

    • Deployment: Guidance on deploying the self-contained ML model package to a SageMaker endpoint within your AWS environment.
    • Optimization: Best practices for maximizing real-time performance and low-latency inference on your chosen instance type.
    • Architecture: Custom architecture strategies tailored to your specific use case, including scaling and endpoint configuration.
    • Troubleshooting: Assistance diagnosing issues with inference requests, audio input formatting, or endpoint behavior.

    Getting Started

    The deployment workflow follows a straightforward pattern: subscribe through this AWS Marketplace listing, deploy the model package to a SageMaker endpoint, send audio input along with the target utterance, and receive a detailed JSON diagnostic response. A free trial is available so you can evaluate scoring quality in your own environment before committing.

    Getting Help

    To request support, send an email to contact@ekinox.io  with a description of your issue or question. Our team reviews all inquiries and responds to work with you on resolution.

    Refunds

    For questions about billing, subscription changes, or refund requests, please contact us at contact@ekinox.io .

    AWS infrastructure support

    AWS Support is a one-on-one, fast-response support channel that is staffed 24x7x365 with experienced and technical support engineers. The service helps customers of all sizes and technical abilities to successfully utilize the products and features provided by Amazon Web Services.

    Similar products

    Customer reviews

    Ratings and reviews

     Info
    0 ratings
    5 star
    4 star
    3 star
    2 star
    1 star
    0%
    0%
    0%
    0%
    0%
    0 reviews
    No customer reviews yet
    Be the first to review this product . We've partnered with PeerSpot to gather customer feedback. You can share your experience by writing or recording a review, or scheduling a call with a PeerSpot analyst.