Machine Learning Engineer: LLM Interpretability & Systems

About this role

Join CTGT as a Senior Machine Learning Engineer to build production systems that make large language models reliable and controllable for enterprise use. You'll work directly with model internals using mechanistic interpretability techniques to operationalize research and enable deterministic AI governance.

What you'll do

• Translate mechanistic interpretability research into production-grade code that improves model behavior
• Develop and implement activation patching, control vectors, and feature extraction techniques for targeted model improvements
• Build evaluation and deployment systems to reliably ship model modifications to enterprise environments
• Design feature-level intervention systems for real-time policy enforcement at inference time
• Work with commercial and open-source model internals across weights, activations, and architectures
• Probe model mechanics to identify fundamental improvements beyond prompt engineering

What they're looking for

• Transformer architectures and PyTorch internals
• Deep learning mathematical foundations
• Model training, fine-tuning, and optimization experience
• Python, Rust, Node/TypeScript programming
• Research paper implementation and evaluation
• Machine learning systems design and deployment
• Problem ownership and debugging complex systems
• Mechanistic interpretability techniques

Benefits

• Competitive base compensation with significant equity in venture-backed company
• Work on core systems with direct real-world impact in high-stakes environments
• High autonomy and trust to form and execute technical opinions
• Access to frontier ML infrastructure and models
• Backing from prestigious investors including Google Gradient Ventures and Y Combinator