Principal ML Engineer: Post-Training to Production

Lila Sciences, Inc.; Cambridge (MA); On-site; USD 252,000 - 336,000; Full time

Job summary

Lila Sciences, Inc. seeks a Principal Machine Learning Engineer in Applied AI to turn advanced AI models into dependable, customer-ready workflows.

The role centers on post-training methods, evaluation loops, and production-oriented ML systems, with emphasis on reliable performance in customer scientific contexts. You will lead post-training efforts using SFT and RL, design evaluation loops, translate customer learnings into model improvements, and mentor engineers.

Qualifications

• Hands-on post-training ML experience with SFT/RL methods.
• Strong Python and PyTorch programming skills.
• Experience debugging complex model behavior and validating with data and logs.

Responsibilities

• Address the gap between model capability and customer scientific workflows.
• Lead post-training efforts using SFT, RL methods (DPO/PPO/GRPO).
• Design evaluation loops to measure model quality and reliability.
• Translate customer feedback into model improvements.
• Collaborate with AI researchers to produce reliable capabilities.
• Integrate model behavior into end-to-end product workflows.
• Mentor engineers and share reusable deployment patterns.

Skills

Python

PyTorch

Post-training ML

RL methods

Model evaluation

Team mentoring

Tools

MoE architectures

Evaluation harnesses

Job description

Lila Sciences, Inc. seeks a Principal Machine Learning Engineer in Applied AI to turn advanced AI models into dependable, customer-ready workflows.

The role centers on post-training methods, evaluation loops, and production-oriented ML systems, with emphasis on reliable performance in customer scientific contexts. You will lead post-training efforts using SFT and RL, design evaluation loops, translate customer learnings into model improvements, and mentor engineers.