Our client, an ambitious AI startup building general-purpose world models for simulation, is seeking a Research Engineer to focus on multimodal foundational models. This individual will work alongside a highly creative team to push the boundaries of interactive, real-time generation. The ideal candidate takes a full-stack approach to research and thrives in a fast-paced environment, taking models all the way from pretraining to production inference.

Role

• Run experiments to teach models new interactive behaviors, such as scene manipulation, action following, and camera control.
• Develop new data strategies, architectural variants, and training techniques while designing rigorous capability evaluations.
• Take features from research prototypes to production by collaborating closely with product and creative teams.

Essential Skills

• 4+ years of experience in machine learning research or engineering.
• Strong familiarity with the architecture, training, and inference of large-scale multimodal generative models.
• Proficiency with major ML frameworks (e.g., PyTorch or JAX) and hands-on experience with distributed training at scale.

Beneficial Skills

• Proven experience building robust data pipelines for both pretraining and post-training (SFT/RL).
• A strong drive to build AI systems that simulate the real world to solve complex problems.

Competitive base salary + equity.

If interested, reply and Goliath Partners will be in touch!