Apply Now

Senior AI/ML Engineer - Speech & Language Assessment (R&D)

Digital Unicorn

DIGITAL UNICORN



📍 Location: 3F, 94 Ho Nghinh Street, An Hai Bac Ward, Son Tra District, Da Nang City

🌐 Website: https://digitalunicorn.fr

📩 Email: hi@digitalunicorn.fr

📞 Phone: 02363 57 57 45

WHO ARE WE?


- Digital Unicorn is a French digital agency with a passionate and ambitious team. We specialize in:


- Mobile Application Development

- Website Development

- AI/Deep Learning

- UX/UI Design


We’re more than just a company — we’re a small hub for those who want to step into an international working environment. Here, you’ll train, work, and build your career foundation, all while growing friendships that last.


At Digital Unicorn:


We share and support each other

We build meaningful products for startups

We create a space where you can enjoy working and growing


WE ARE HIRING: 

Senior AI/ML Engineer - Speech & Language Assessment (R&D)


Level: [Senior]

 

About the Project

We are building BriLanguage, an AI-powered French language proficiency assessment platform from the ground up. Similar to Elsa Speak but designed specifically for French, our solution will provide comprehensive CEFR level evaluation (A1-C2) through advanced speech-to-text analysis, including detailed phonetic assessment, accent detection, and linguistic pattern recognition.

We are seeking an AI/ML expert specializing in open-source LLMs and speech processing to lead the R&D process for this greenfield project. This is a critical role that requires not just technical execution, but strategic thinking to define project scope, establish realistic timelines, and ensure successful delivery within budget constraints.

This is primarily a technical leadership and planning role - we need someone who can architect the solution, assess feasibility, and guide our development roadmap.

Key Responsibilities

Strategic Planning (Primary Focus)

  • Define technical scope: Break down the complete project into achievable milestones
  • Timeline estimation: Create realistic development roadmap with clear deliverables
  • Resource planning: Determine required compute resources, data needs, and team composition
  • Budget assessment: Validate if project is feasible within initial budget or propose necessary adjustments
  • Risk analysis: Identify technical challenges and propose mitigation strategies

Technical Leadership

  • Design the end-to-end AI architecture for speech-based CEFR assessment
  • Select and justify appropriate open-source models and frameworks
  • Define data collection and annotation requirements
  • Establish evaluation metrics and validation methodology
  • Guide technical decision-making throughout R&D phase

Hands-on Development (As Needed)

  • Prototype core components of the assessment pipeline
  • Fine-tune speech recognition models for French phonetics
  • Develop CEFR scoring algorithms and rubrics
  • Implement proof-of-concept demonstrations

Required Qualifications

Must-Have Experience

1. Speech Processing & LLMs

  • Deep expertise with open-source LLM frameworks (Hugging Face, PyTorch/TensorFlow)
  • Proven experience in speech recognition and ASR systems (Whisper, Wav2Vec2, HuBERT, Kaldi, etc.)
  • Hands-on work with speech-to-text pipelines in production environments

2. Language Assessment Background

  • Direct experience with language proficiency assessment tools (Elsa Speak, Duolingo English Test, or similar platforms strongly preferred)
  • Understanding of CEFR framework and automated scoring methodologies
  • Experience building or working with pronunciation/phonetic analysis systems

3. R&D Leadership

  • Track record of leading AI/ML projects from concept to production
  • Experience scoping complex ML projects with timeline and resource estimation
  • Ability to make build-vs-buy decisions and justify technical tradeoffs

4. Technical Skills

  • Proficiency in Python and ML frameworks (PyTorch, TensorFlow, scikit-learn)
  • Experience with speech processing libraries (librosa, Montreal Forced Aligner, ESPnet)
  • Knowledge of MLOps practices and model deployment

Highly Desirable

  • French language expertise or experience with French phonetics/linguistics
  • Background in computational linguistics or phonetics research
  • Experience with multilingual speech models
  • Published work in speech recognition, NLP, or language assessment
  • Previous work building educational technology or language learning applications
  • Experience managing technical budgets and vendor relationships

What We Need From You

Initial Deliverables (First 2-4 Weeks)

  1. Technical feasibility study: Comprehensive analysis of approach, risks, and alternatives
  2. Detailed project scope: Feature breakdown with technical specifications
  3. Timeline & milestones: Realistic development roadmap with dependencies
  4. Resource requirements:
    • Data needs (hours of labeled speech, speaker diversity)
    • Compute infrastructure (GPU/cloud requirements)
    • Team composition and skills needed
  6. Budget analysis: Detailed cost breakdown and funding recommendations

Ongoing

  • Technical architecture documentation
  • Regular progress assessments and course corrections
  • Prototyping and validation of core components
  • Knowledge transfer and team mentorship

Application Requirements

Please include:

  1. Portfolio/GitHub: Links to relevant projects, especially speech/NLP work
  2. Relevant experience: Specific examples of language assessment or speech recognition projects
  3. Brief proposal (1-2 pages):
    • Your initial thoughts on technical approach
    • High-level timeline estimate for MVP
    • Key risks you foresee



Apply to join Digital Unicorn by sending your CV to hi@digitalunicorn.fr with the subject line "[Job title]_[FullName]"


Thank you!

--Digital Unicorn--
Apply Now