Member of Technical Staff (Infrastructure Engineer, Training and Inference Systems)

About the Company

Inherent is a frontier AI lab dedicated to developing AI that explores "unknown unknowns" to uncover paradigm-shifting research contributions. We are building a new research factory where AI agents accelerate their own creation through recursive self-improvement.

Responsibilities

• Contribute to training and inference infrastructure end-to-end
• Build systems for large-scale experiments including distributed job orchestration, data pipelines, and evals
• Implement and harden RL and post-training libraries for models
• Close recursive loops to enable AI agents to drive their own training and infrastructure
• Collaborate closely with the core research team on their primary systems

Requirements

• Experience with high-performance, large-scale distributed systems, preferably for LLM workloads
• Proficiency in Python and PyTorch or JAX
• Hands-on experience with large-scale LLM training or inference technologies (e.g., SGLang, vLLM, verl, Megatron, OpenRLHF)
• Strong technical judgment regarding building vs. buying solutions
• Alignment with an agent-centric approach to company building

Preferred Qualifications

• Experience with systems programming languages like Rust or C++
• Experience writing and profiling CUDA kernels
• A track record of building reliable research tools