Member of Technical Staff - Applied Machine Learning Scientist

Liquid AI, Inc.; San Francisco (CA); On-site; USD 120,000 - 180,000; Full time

Job summary

An innovative company is seeking a Machine Learning Architect to lead impactful experimentation at the intersection of foundational model research and real-world applications. In this hands-on role, you'll manage the entire lifecycle of experiments, ensuring insights are actionable and aligned with product goals. Collaborating closely with cross-functional teams, you'll design tailored training setups and evaluation methods, driving measurable improvements in model performance. This position offers a unique opportunity to shape the future of AI solutions in a dynamic, high-autonomy environment.

Qualifications

• Experience with foundation models beyond just LLMs.
• Deep experience with the PyTorch ecosystem.

Responsibilities

• Own the end-to-end experimental process from hypothesis to results.
• Design and run structured experiments to evaluate model performance.

Skills

PyTorch

Machine Learning

Experiment Design

Model Evaluation

Customer Feedback Integration

Job description

Our goal at Liquid is to build the most capable AI systems to solve problems at every scale, such that users can build, access, and control their AI solutions. This is to ensure that AI will get meaningfully, reliably and efficiently integrated at all enterprises. Long term, Liquid will create and deploy frontier-AI-powered solutions that are available to everyone.

We're looking for a Machine Learning Architect to lead high-leverage experimentation at the intersection of foundational model research and real-world customer impact. This is a hands-on, research-driven role where you'll own the end-to-end lifecycle of designing, running, and analyzing experiments that push forward our hybrid model systems.

Your work will directly shape how our models perform in customer contexts, with an emphasis on measurable impact over theoretical gains. You’ll collaborate closely with infra, modeling, and customer teams to ensure that experimental insights are actionable and aligned with product goals.

You'll be a great fit if

• You thrive in high-autonomy, high-context environments and know how to turn vague questions into concrete experiments.
• You’ve worked on foundation models beyond just LLMs, and you understand the nuances of designing for real-world signals and feedback.
• You’re comfortable creating new training setups, loss functions, or evaluation methods tailored to customer-specific metrics.
• You have deep experience with the PyTorch ecosystem (including distributed training and third-party libraries), and can move quickly from prototype to production-scale experiments.
• You’re energized by tight feedback loops with customers and believe that experimentation should be aligned with product objectives.
• You can think at multiple altitudes—from quick-turn tests to longer-term architecture bets—and know when to scale each.

What you'll actually do

• Own the end-to-end experimental process: from hypothesis generation to results analysis to iteration.
• Design and run structured experiments to evaluate model performance across customer-specific metrics.
• Develop tooling and workflows that let the team rapidly test hypotheses and scale promising directions.
• Work closely with research, infra, and customer teams to prioritize experimental goals and interpret results.
• Translate customer needs into actionable model improvements via principled experimentation.
• Contribute to building a culture of fast, reproducible, and product-aligned model research.