Member of Technical Staff - Research Engineer, Omni

Liquid AI • Boston, MA

4d • Hybrid

About The Position

Liquid AI is hiring a Research Engineer for its Omni (Multimodal) team in Research & Engineering. Omni brings the company’s vision and audio work together in a single multimodal model, with open research directions including multimodal mixture-of-experts (MoE) and token-efficient encoders. This is an end-to-end model training role spanning research, data, pre- and post-training, evaluation, inference, and deployment. The role works directly with the Omni lead and vision and audio teams, with close involvement from the Chief Science Officer. The team is small, has recently shipped a release, and is consolidating its training infrastructure.

The Research Engineer will run multimodal training experiments and help carry research results into future Omni releases. The role also contributes to multimodal research, data pipelines and curation, maintainable training infrastructure, distributed-training decisions, and inference and deployment. Liquid AI is open to experienced engineers and strong early-career candidates.

Liquid AI builds general-purpose AI systems designed to run efficiently across deployment targets, from data-center accelerators to on-device hardware, with a focus on low latency, minimal memory usage, privacy, and reliability. The company partners with enterprises in consumer electronics, automotive, life sciences, and financial services.

Requirements

• Hands-on experience training machine-learning models in at least one non-text modality; vision is the most likely path, while audio can be learned on the job.
• Foundational understanding of how multimodal training works below the research layer, and ability to explain it based on past work.
• Ability to work between research and engineering by running experiments and hardening the code behind them.
• Ability and willingness to take training infrastructure and inference from experiment code to maintained, production-ready systems and own that process.

Nice To Haves

• Audio modality and speech-modeling experience.
• Prior inference or deployment experience, including on-device inference.
• Exposure to multimodal MoE or encoder token-efficiency research.
• Publication record or academic research background; understanding what you built is what matters.
• Enjoyment of infrastructure and inference work.
• Experience consolidating or rebuilding a training stack.
• Opinions on libraries and distributed-training paradigms.

Responsibilities

• Run multimodal training experiments end to end, including pre-training and post-training, and incorporate results into the next Omni release.
• Contribute to open research directions such as multimodal MoE and token efficiency for multimodal encoders.
• Contribute to multimodal data pipelines, curation, and the data decisions that feed each release.
• Harden research-grade experiment code into maintainable infrastructure in the team repository.
• Help consolidate Omni’s training infrastructure, including establishing a unified codebase and library and making and defending distributed-training choices.
• Own inference and deployment for Omni models.
• Work across the vision and audio teams on the connections between data, infrastructure, training, and evaluations.
• Help expand the team’s research bandwidth and drive new capabilities, such as multimodal MoE, into shipped releases.

Benefits

• Competitive base salary plus equity.
• Unlimited PTO.
• Company-wide Refill Days and long weekends when the whole company rests together.
• Comprehensive health coverage, with details varying by location.
• Paid leave for pregnancy, parenting, medical needs, and caring for family.

What This Job Offers

• Job Type: Full-time
• Career Level: Entry Level
• Education Level: No Education Listed