Member of Technical Staff – Research Engineer, Omni

Liquid AI

. Run multimodal training experiments end to end, including pre- and post-training .

Posted 11/11/2026full-timeBoston • Massachusetts • United StatesLeadWebsite

Core Competencies

Role fit

Core Competencies

Use this summary to align your resume positioning with the role.

Demonstrates expertise in multimodal training, including the ability to develop and maintain production-ready systems for training and inference. Strong collaboration skills across teams to drive new capabilities and consolidate training infrastructure.

Highest-signal resume keywords

Multimodal Training ExperienceInfrastructure DevelopmentInference and Deployment OwnershipData Pipeline DevelopmentDistributed-Training Knowledge

ATS Keywords

Tailor your resume

Applicant Tracking System Keywords

Tip: use these terms in your resume and cover letter to boost ATS matches.

Hard Skills

Machine Learning Model TrainingMultimodal Data ProcessingExperiment Code MaintenanceProduction-Ready Systems DevelopmentToken Efficiency Research

Soft Skills

CollaborationCommunicationProblem-Solving

Tools & Technologies

Training InfrastructureOmni ModelsData Curation Tools

Industry Keywords

Multimodal MoEAudio ModalitySpeech ModelingOn-Device Inference

About the role

Key responsibilities & impact

Run multimodal training experiments end to end, including pre- and post-training
Fold experiment results into the next Omni release
Contribute to research directions such as multimodal MoE and token efficiency for multimodal encoders
Develop multimodal data pipelines and curate data for releases
Make data decisions that feed each release
Harden research-grade experiment code into maintainable infrastructure
Consolidate Omni's training infrastructure into one codebase and one library
Make and defend distributed-training choices
Own inference and deployment for Omni models
Collaborate across vision and audio teams on data, infrastructure, training, and evaluations
Drive new capabilities into shipped releases
Run Omni model training end to end, including infrastructure, inference, and half of the data work

Requirements

What you’ll need

Hands-on experience training ML models in at least one non-text modality
Foundational understanding of multimodal training
Ability to explain multimodal training from personal past work
Comfortable working between research and engineering
Ability to take training infrastructure and inference from experiment code to maintained, production-ready systems
Willingness to own the path from experimentation to production
Strength in infrastructure and inference
Experience consolidating or rebuilding a training stack preferred as an area of strength
Opinions on libraries and distributed paradigms
Audio modality and speech modeling experience helpful but not required
Prior inference or deployment experience, including on-device inference, helpful but not required
Exposure to multimodal MoE or encoder token-efficiency research helpful but not required
Publication record or academic research background helpful but not required

Benefits

Comp & perks

Competitive base salary plus equity
Unlimited PTO
Company-wide Refill Days
Long weekends when the whole company rests together
Comprehensive health coverage
Paid leave for pregnancy, parenting, medical needs, and caring for family
End-to-end ownership of a shipped multimodal model
Research time that grows as the stack stabilizes