Senior / Staff Machine Learning Engineer, Applied AI
Lila Sciences2 months ago
Cambridge, MA, USA or San Francisco, CA, USASenior / Staff+
H1B Sponsor
Responsibilities
- Close the last-mile gap between Lila AI model capabilities and customer-specific scientific workflows.
- Build evaluation loops that measure model quality, reliability, and customer fit.
- Design experiments to improve model performance across applied customer use cases.
- Feed customer learnings, data signals, and evaluation results into AI model improvement cycles.
- Partner with AI researchers to translate model improvements into usable capabilities.
- Work with software teams to integrate model behavior into end-to-end product workflows.
- Debug model failures using traces, evaluations, customer context, and scientific feedback.
- Build reusable tooling for model adaptation, evaluation, and deployment workflows.
Requirements
- Strong experience building, training, adapting, or evaluating machine learning models.
- Strong software engineering skills in Python and modern machine learning frameworks such as PyTorch, JAX, or TensorFlow.
- Experience with distributed machine learning training frameworks including Megatron-LM, TorchTitan, DeepSpeed, or Ray.
- Experience designing experiments, evaluation metrics, or test sets for model performance.
- Ability to debug model behavior using data, traces, logs, and qualitative feedback.
- Experience collaborating across research and engineering teams to move machine learning capabilities into usable systems.
- Familiarity with large language models, multimodal models, or agentic AI systems.
- Clear communication skills for translating customer needs into technical model improvements.
- Preferred experience adapting models for customer-facing or production workflows; working with scientific, technical, or data-intensive use cases; building evaluation harnesses, model monitoring, or quality dashboards; using retrieval-augmented generation, tool use, or agentic workflows; reinforcement learning post-training such as RLHF, GRPO, or tool-augmented RL; training mixture-of-experts architectures; and collaborating with product or customer-facing teams.
Categories
About Lila Sciences
Lila Sciences is the world’s first scientific superintelligence platform and autonomous lab for life, chemistry, and materials science. We are building the foundation to apply AI to every aspect of the scientific method, enabling scientists to bring forth solutions in human health and sustainability at a pace and scale never experienced before.