Member of Technical Staff: Research Engineer, Post-Training

Essential AI's mission is to deepen the partnership between humans and computers, unlocking collaborative capabilities that far exceed what could be achieved today. We believe that building delightful end-user experiences requires innovating across the stack - from the UX all the way down to models that achieve the best user value per FLOP.

The Research Engineer, Post-Training will be responsible for developing and implementing techniques to optimize and fine-tune models after the initial training process, with the goal of improving performance, robustness, and efficiency. You will work cross-functionally to identify areas for post-training optimization and measure the impact on model performance. You will also be responsible for benchmarking and evaluating post-training techniques on a variety of datasets and model architectures.

Key responsibilities include:

• Leading or being a core contributor to research bets that advance the real-world capabilities of our models
• Collaborating closely with product teams to close the loop between research and product
• Developing and implementing novel post-training techniques to optimize machine learning models
• Benchmarking and evaluating post-training techniques on various datasets and model architectures
• Implementing post-training models and algorithms, optimizing for performance and scalability

We are looking for candidates with:

• Research experience focusing on post-training and optimizing large language models
• Strong ML fundamentals and first principles thinking
• Experience in developing new methods or improving existing techniques in ML
• Proficiency in programming languages such as Python, C++, or Java
• Familiarity with ML deployment and orchestration

We are based in-person in SF and fully onsite 5 days a week, offering relocation assistance to new employees. Join us in building a world-class multi-disciplinary team excited to solve hard real-world AI problems.