Research Engineer

Harnham; California; On-site; USD 150,000 - 230,000; Full time

Job summary

Harnham in California is seeking a data-focused ML engineer to own data powering next-generation AI systems. You will design multimodal datasets, run controlled experiments, and build scalable data pipelines for synthetic data generation and quality control.

You will define evaluation benchmarks and collaborate with cross-functional teams to translate product goals into data strategies, leveraging PyTorch or JAX and distributed compute tools.

Qualifications

• 4+ years of experience in machine learning, ideally with a data-centric focus.
• Experience working with large multimodal datasets and generative models.
• Strong intuition for how data quality and composition impact model behavior.
• Experience across the full ML lifecycle, from data to training to evaluation.

Responsibilities

• Design multimodal, multitask datasets to unlock new model capabilities.
• Run controlled experiments to understand how data impacts model performance.
• Build and scale pipelines for synthetic data generation, filtering, and quality control.
• Define evaluation frameworks and benchmarks to measure real-world model improvement.
• Partner with cross-functional teams to translate product goals into data strategies.

Skills

Multimodal datasets

Data-centric ML

PyTorch

JAX

Ray

Kubernetes

Tools

PyTorch

JAX

Ray

Kubernetes

Job description

We're partnered with a well-funded AI research company focused on building next-generation multimodal models for media and interactive experiences. Their work spans cutting-edge generative systems and is increasingly moving toward real-time, interactive environments, pushing beyond static outputs into dynamic, AI-driven applications.

This is a high-impact, research-meets-engineering role focused on the data that powers advanced AI systems. You'll own how models learn, designing datasets, running experiments, and building data pipelines that directly shape model capabilities across a wide range of applications.

What You'll Do

• Design multimodal, multitask datasets to unlock new model capabilities
• Run controlled experiments to understand how data impacts model performance
• Build and scale pipelines for synthetic data generation, filtering, and quality control
• Define evaluation frameworks and benchmarks to measure real-world model improvement
• Partner with cross-functional teams to translate product goals into data strategies

Requirements

• 4+ years of experience in machine learning, ideally with a data-centric focus
• Experience working with large multimodal datasets and generative models
• Strong intuition for how data quality and composition impact model behavior
• Experience across the full ML lifecycle, from data to training to evaluation
• Proficiency with ML frameworks such as PyTorch or JAX
• Experience with distributed systems or compute tools (e.g., Ray, Kubernetes)
• Strong interest in advancing next-generation AI systems

Nice to Have

• Experience with synthetic data generation or data curation at scale
• Background working on multimodal or video-based models
• Exposure to evaluation and benchmarking for generative systems

If you're interested in shaping the data that defines what AI models can learn and do, this is a unique opportunity to work at the forefront of multimodal AI.