Machine Learning Engineer – TTS

This is a fully remote position, open to applicants in Spain.

📋 Description

• Expand TTS capabilities to new languages through data gathering and curation, text normalization/G2P, and model adaptation.

• Transition new TTS architectures from research phases into production environments.

• Enhance model robustness, latency, and quality in collaboration with MLOps during deployment.

• Elevate existing models through post-training, data curation, architectural enhancements, and novel training methodologies.

• Construct efficient audio data pipelines, covering speaker recording, preprocessing, and model training.

• Develop thorough benchmarks and evaluations to assess whether new model iterations demonstrate improvement and the reasons behind it.

• Collaborate within a compact TTS team comprising 2–3 ML engineers, with support from MLOps and data engineering teams.

⛳️ Requirements

• Over 3 years of experience in speech technology, specifically with practical TTS experience.

• Direct experience with modern LLM-based TTS methodologies (e.g., VALL-E, CosyVoice, XTTS).

• Proven ability to make TTS models production-ready, ensuring stability, robustness, and inference efficiency at the model level.

• Expertise in deep learning frameworks, particularly PyTorch.

• Comprehension of contemporary ML principles: Transformers, diffusion/flow matching, neural audio codecs, and vocoders.

• Knowledge of streaming TTS limitations (latency, chunked/incremental generation) and their impact on model architecture decisions.

• Familiarity with audio signal processing and the linguistic aspects of TTS (phonetics, text normalization).

• Basic experience with speech data processing (segmentation, alignment, quality filtering).

• Awareness of TTS evaluation techniques (MOS, SBS).

• Strong proficiency in Python and solid software engineering principles.

• Demonstrated ability to independently investigate new concepts and enhance existing ML systems.

• Bachelor’s degree or higher in Computer Science or a related field, or equivalent experience.

• Proficiency in English at a B2 level or higher.

• Experience with low-resource languages, including dataset creation and augmentation.

• Experience in designing recording scripts and specifications for TTS voice datasets.

• Familiarity with MLOps/DataOps tools such as MLflow, DVC, or similar experiment tracking platforms.

• Proficiency in C++.

• Publications in speech or ML-related venues.

🏝️ Benefits

• 21 vacation days plus public holidays and 5 sick days.

• Private English lessons through Preply.

• Fully remote work available across Europe.

• Access to a cutting-edge tech stack: Speech Technologies, NLP, Generative AI (LLMs, diffusion models), and a voice-first agentic architecture with a focus on privacy and on-premises deployment.

• High engineering standards and genuine ownership of projects.

• Opportunities for rapid career advancement.

• A startup environment with the stability of an enterprise — real clients, real revenue, and minimal bureaucracy.