Software Engineer, Applied AI Infrastructure

Nuro•Mountain View, CA

About The Position

Nuro is seeking a Software Engineer for Applied AI Infrastructure to build the platform that lets AI agents operate autonomously within Nuro's engineering organization. The goal is to amplify the output of every engineer and researcher by 100x, focusing on creating a trustworthy and rigorous closed-loop evaluation system for AI work. This role involves building a substantial agent system in production, making it trustworthy, autonomous, and significantly more capable. The core responsibilities include developing closed-loop evaluation, the agent platform, and autoresearch infrastructure. The work will involve building the measurement layer for agent output, transitioning the autoresearch loop to unattended operation, and designing the isolation and permissioning model for agents acting on production systems. The role offers direct access to compute, systems being automated, and interaction with engineering leadership and the CEO.

Requirements

3+ years of software engineering experience (or 2+ with a Master's) in computer science, engineering, or equivalent practical experience.
Deep, current understanding of LLM research, including model training from scratch (data, tokenization, architecture, pretraining dynamics) and the post-training stack (supervised fine-tuning, preference optimization, RL).
Ability to reason about how training decisions affect model behavior and follow relevant literature.
Understanding of inference under the hood, including attention and KV-cache behavior, batching and scheduling, quantization, speculative decoding, prefix caching, context handling, and their trade-offs.
Experience building and operating LLM-based agent systems in production (tool use, orchestration, sandboxing, retrieval, memory).
Strong backend and distributed systems background at scale (cloud infrastructure, service design, storage, queuing).
Strong programming skills in Python.
Strong opinions and experience with evaluation, including arguing about the validity of benchmarks.
Ability to work end-to-end without pre-decomposed problems.
Hands-on post-training or fine-tuning experience (SFT, preference optimization, RL, distillation), including evaluation.
Experience with ML training or research infrastructure (experiment orchestration, evaluation pipelines, hyperparameter search, data pipelines).
Experience running inference serving, cost, or capacity at meaningful scale.
Familiarity with agent architecture patterns (planning, reflection, long-horizon memory, multi-agent coordination).
Experience with open tool-integration protocols, plugin or skill frameworks, and model-routing or gateway layers.
Background in developer experience or platform engineering, observability, or security isolation.

Nice To Haves

Staff-level candidates should bring correspondingly deeper scope and ownership.
Go, C++, or Rust experience in addition to Python.

Responsibilities

Build the closed-loop measurement layer that determines whether agent output is accepted, reverted, or overridden, and use it to guide the expansion or retraction of autonomy.
Take the autoresearch loop from assisted to unattended for a bounded class of experiments, including the necessary evaluation and confidence machinery.
Design the isolation and permissioning model that allows agents to act on production repositories and infrastructure with an auditable record of their actions and rationale.
Develop and operate LLM-based agent systems in production, including tool use, orchestration, sandboxing, retrieval, and memory.
Build and operate backend and distributed systems at scale, including cloud infrastructure, service design, storage, and queuing.
Implement agent-powered tooling across the engineering lifecycle, such as code generation, review, debugging, test and CI failure attribution, knowledge retrieval, and triage.
Potentially work on post-training internal models where justified by internal workload.