At Liquid, we’re not just building AI models—we’re redefining the architecture of intelligence itself. Spun out of MIT, our mission is to build efficient AI systems at every scale. Our Liquid Foundation Models (LFMs) operate where others can’t: on-device, at the edge, under real-time constraints. We’re not iterating on old ideas—we’re architecting what comes next.
We believe great talent powers great technology. The Liquid team is a community of world-class engineers, researchers, and builders creating the next generation of AI. Whether you're helping shape model architectures, scaling our dev platforms, or enabling enterprise deployments—your work will directly shape the frontier of intelligent systems.
This Role Is For You If:You like building LLM pipelines and agents for diverse use cases, and enjoy catching and fixing edge cases where LLMs may fail
You’re a native Japanese speaker and want to further improve LLM capabilities in Japanese
You’re motivated by the challenge of adapting foundation models to new languages, cultures, and enterprise workflows
Deep understanding of the Japanese model evaluation landscape and familiarity with Japanese pre-training data sources
Experience using modeling and inference tools such as Huggingface inference, vLLM, and cloud APIs
Identify, collect, and curate diverse high-quality Japanese text, audio, and multimodal datasets
Design methods to synthetically generate or augment Japanese training data when needed
Ensure datasets meet enterprise-grade quality, coverage, and compliance requirements
Train and fine-tune language and vision models to achieve state-of-the-art performance for Japanese enterprise use cases
Adapt existing LFMs for Japanese language, culture, and enterprise-specific workflows
Implement evaluation frameworks to benchmark model quality on Japanese datasets
Design evaluation datasets and metrics for Japanese enterprise applications
Conduct thorough error analysis and iteratively improve model performance
Ensure robustness, fairness, and reliability in Japanese-language outputs
Hands-on experience with state-of-the-art technology at a leading AI company
The opportunity to directly shape foundation model performance in one of the world’s most complex and nuanced languages
A collaborative, fast-paced environment where your work drives the next generation of LFMs
Spun out of MIT CSAIL, we’re a foundation model company headquartered in Boston. Our mission is to build capable and efficient general-purpose AI systems at every scale—from phones and vehicles to enterprise servers and embedded chips. Our models are designed to run where others stall: on CPUs, with low latency, minimal memory, and maximum reliability. We’re already partnering with global enterprises across consumer electronics, automotive, life sciences, and financial services. And we’re just getting started.
Top Skills
What We Do
Our mission is to build capable and efficient general-purpose AI systems at every scale
