Research Engineer – Large Language Models (Foundational Model from Scratch)

Bay Area, CA

A deep-tech startup on a mission to build a foundational AI model from scratch. Our team combines world-class research, cutting-edge engineering, and a bold vision: to create AI systems that are causal, interpretable, and scalable. Backed by top-tier investors and pioneering talent in AI, we’re tackling some of the hardest problems in modern machine learning.

We are looking for a Research Engineer who has hands-on experience contributing to large language models (LLMs) — whether that’s pretraining, post-training, or fine-tuning. You will join our team building a foundational model from scratch, shaping its architecture, training pipelines, and capabilities. This is a rare opportunity to work at the forefront of AI research and engineering in the Bay Area.

What You’ll Do:

* Lead and contribute to core development of a foundational language model, including model architecture, training strategies, and evaluation.

* Implement, optimize, and scale pretraining, fine-tuning, and post-training pipelines for large-scale models.

* Collaborate closely with research scientists and ML engineers to push the boundaries of LLM capabilities.

* Conduct experiments to improve model performance, generalization, and safety.

* Build reusable infrastructure and tools for training, evaluation, and deployment of next-generation LLMs.

What We’re Looking For:

* Strong hands-on experience with large language models, ideally with contributions to pretraining, fine-tuning, or RLHF.

* Expertise in PyTorch or JAX, distributed training, and large-scale ML pipelines.

* Solid understanding of modern LLM architectures and training techniques.

* Ability to translate research insights into scalable engineering solutions.

* Passion for building models from the ground up and contributing to a high-performing, collaborative team.

Apply now for a confidential chat