Liquid AI

Build and extend the inference engine and benchmarking infrastructure that runs AI models across data-center and on-device hardware. This role combines performance-sensitive C++ and Python development, model porting, benchmark design, and technical validation with external partners.

Join Liquid AI as an ML Scientist to enhance foundation models for the Japanese market, focusing on post-training strategies and experimental methodologies.

Deploy advanced language and multimodal models into reliable, efficient production systems for customers in Japan. This highly autonomous role combines hands-on ML engineering, inference optimization, evaluation, and technical customer collaboration.

Own and improve the GPU cluster infrastructure powering Liquid AI’s foundation model training and research. You’ll combine hands-on software engineering, operational response, and automation to make compute platforms more reliable and efficient.

Build Liquid AI’s solutions architecture function while taking AI customer engagements from technical discovery through go-live. You’ll create demos and proofs of concept, translate efficient models into customer outcomes, and shape product direction through field feedback.

Own end-to-end applied post-training for vision-language models, combining enterprise customer delivery with hands-on multimodal model development. You will curate data, fine-tune and align VLMs, design evaluations, and help advance Liquid AI’s core post-training stack.

Own end-to-end applied ML engagements that adapt large-scale sequential recommendation models for enterprise customers. You will build data and evaluation workflows, fine-tune models, and deliver production-focused personalization and ranking systems.

Own the applied post-training work that adapts Liquid AI’s audio language models for enterprise voice-driven function calling. You’ll build data and evaluation pipelines, fine-tune and align models, and ship real-time, on-device audio AI solutions.