AI Engineer (Whisper,MLOps,Fine Tuning), FusionHit, CR

FusionHit is seeking an experienced AI Engineer to join our dynamic team and take ownership of a high-impact project. This role involves self-hosting and fine-tuning OpenAI’s Whisper model (ideally WhisperX) for transcription and ambient listening use cases. You’ll also establish a robust MLOps pipeline for model retraining and deployment in a production environment.

The ideal candidate is a hands-on ML practitioner with a deep understanding of speech-to-text systems and cloud infrastructure. This is a mission-critical role with high visibility, where you'll help deliver a scalable, production-grade AI solution by year-end.

Location: Must reside and have work authorization in Latin America. This is a freelancing opportunity.
Availability: Must be available to work with significant overlap with Mountain Standard Time (MST).

The Ideal Candidate Has:

• BS/MS in Computer Science, Machine Learning, or related field with 5+ years of experience in AI/ML engineering.
• Deep experience with speech-to-text models such as Whisper or WhisperX.
• Proven expertise in fine-tuning ML models with labeled datasets.
• Strong experience in MLOps using tools like MLflow, Kubeflow, or similar frameworks.
• Hands-on experience deploying models on Azure (self-hosted, not managed services).
• Proficiency in Python and ML libraries like PyTorch or TensorFlow.
• Experience working with audio datasets and preprocessing techniques.
• Familiarity with prompt engineering related to speech-based AI solutions.
• Excellent communication skills in English (C1 preferred, strong B2 may be considered).

Key Responsibilities:

• Fine-tune Whisper/WhisperX models for transcription and ambient listening tasks.
• Deploy self-hosted Whisper models on Azure cloud infrastructure.
• Design and implement an MLOps pipeline to support iterative training and deployment.
• Ensure high data quality using existing audio + transcript datasets.
• Collaborate on prompt engineering strategies for speech recognition improvements.
• Deliver a production-ready model before year-end.