Member of Technical Staff - Multi-Modal Audio

About the role

Liquid AI is building frontier speech‑language models that combine speech‑to‑text, text‑to‑speech and speech‑to‑speech in a single architecture. As a Member of Technical Staff on the Audio team, you will develop production‑grade audio models that run on‑device under real‑time constraints and ship to customers.

Key responsibilities

• Design, build, and scale data pipelines for audio model training, including preprocessing, augmentation, and quality filtering.
• Develop and maintain evaluation systems that benchmark multimodal performance on internal and public datasets.
• Fine‑tune and adapt audio models for specific customer use cases, managing the full delivery lifecycle from requirements to deployment.
• Contribute production‑ready code to the core audio repository and collaborate with infrastructure and research teams.
• Support experimentation on real hardware, balancing customer work with core development priorities.

Required profile

• Strong programming fundamentals with a track record of writing clean, maintainable, production‑grade code.
• Experience building and shipping production ML systems beyond model training, such as data pipelines, evaluation frameworks, and serving infrastructure.
• Ability to own end‑to‑end outcomes, from data ingestion to customer deployment.
• Comfort working under on‑device, low‑latency, memory‑limited constraints.

Required skills

• Proficiency in PyTorch.
• Familiarity with distributed training frameworks such as DeepSpeed or FSDP.