Backend Engineer, LLM Agent Platform

Anduril Industries; Washington (District of Columbia); Sur place; USD 220 000 - 292 000; Plein temps

Avantages offerts par ce poste

Equity grants

Competitive benefits package

Performance incentives

Résumé du poste

Anduril Industries is seeking a backend software engineer to own and evolve the internal LLM agent framework within the Frontier AI team. You will design, implement, and improve the platform that enables teams to build, evaluate, and deploy reliable LLM agents in mission-critical environments.

The role sits at the intersection of backend infrastructure, AI, agent architecture, model post-training, and evaluation tooling.

Qualifications

• Production-grade backend platforms experience.
• Designing LLM agent frameworks and evaluation methods.
• Experience with model post-training workflows (SFT, RL).
• Reliability, observability, and safety mindset for high-stakes apps.
• Ability to translate ambiguous product needs into reusable platform capabilities.

Responsabilités

• Own and evolve Anduril’s internal LLM agent framework, including core abstractions and runtime architecture.
• Provide framework capabilities and architectural guidance to multiple business lines.
• Collaborate with ML teams on post-training workflows (SFT, RL, environment-driven updates).
• Design tooling for agent patterns (tool calling, memory, planning, subagents, agent graphs).
• Define evaluation suites for task success, tool calls, trajectory quality, robustness and deployment readiness.
• Stay current on agent architecture trends and pragmatic internal adoption decisions.

Connaissances

Backend framework design

LLM agent framework

Model post-training workflows

Reliability & observability

Cross-functional collaboration

Outils

Kubernetes

Docker

Description du poste

Anduril Industries is seeking a backend software engineer to own and evolve the internal LLM agent framework within the Frontier AI team. You will design, implement, and improve the platform that enables teams to build, evaluate, and deploy reliable LLM agents in mission-critical environments.

The role sits at the intersection of backend infrastructure, AI, agent architecture, model post-training, and evaluation tooling.