Responsibilities:

• Lead the design, development, and optimization of ML pipelines for solutions based on STT, TTS, and LLMs.
• Fine-tune speech recognition and language models for low-resource languages and dialects.
• Collaborate with the data annotation team to ensure data quality for model training.
• Transform ML models and prototypes into production-ready pipelines.
• Optimize model performance to achieve accuracy, speed, and efficiency.
• Automate key stages of CI/CD pipelines for deploying ML models.
• Mentor junior machine learning engineers and guide the team on best practices for coding, model management, and MLOps.
• Document processes, write specifications, and develop user guides for ML models and pipelines.
• Ensure compliance with data privacy and model security standards.

Requirements:

• 5+ years of experience in machine learning or an MLE role.
• Proven experience in STT, TTS, or NLP projects.
• Deep knowledge of LLMs and model tuning for low-resource languages.
• Proficiency in Python ML libraries (TensorFlow, PyTorch, HuggingFace, etc.).
• Experience integrating ML models into complex, data-driven systems.
• Knowledge of cloud providers (AWS, GCP, Azure) and MLOps platforms (AWS SageMaker, MLFlow, Kubeflow, etc.).
• Experience managing data annotation pipelines and ensuring data quality.
• Strong communication and leadership skills.

Bonus Points:

• Experience with ASR frameworks (Wav2vec, Kaldi, SpeechBrain, Whisper, etc.).
• Competency in the Apache Airflow ecosystem.
• Familiarity with TTS frameworks (VITS, XTTS, Coqui TTS, Tacotron, etc.).
• Experience in multilingual NLP (tuning multilingual models).
• Knowledge of dialect recognition technologies.
• Competency in the Apache Spark ecosystem.
• Expertise in LLM/Generative AI models.