AI Tooling and Infrastructure Engineer

About the role

Hewlett Packard Enterprise's Networking Training and Documentation organization is building the infrastructure that lets its team use AI to create and maintain product training and documentation. This role designs, builds and operates the pipelines, services and tooling that integrate LLMs into internal tools and customer-facing products, using vendor APIs and/or self-hosted tooling. It is a hybrid position, with about two days per week in an HPE office.

Key responsibilities

• Design, build and operate infrastructure that integrates LLMs into internal tools and customer-facing products.
• Evaluate, implement and maintain third-party AI tooling and platforms, owning the vendor relationship and integration where applicable.
• Build homegrown tooling such as pipelines, orchestration layers, retrieval systems, evaluation harnesses and monitoring when off-the-shelf solutions do not fit.
• Own operational reliability: uptime, latency, cost and observability of AI-powered systems.
• Implement guardrails around cost, rate limits, prompt and version management, and failure handling.
• Work with documentation and training stakeholders to turn use cases into maintainable systems.
• Set up monitoring, logging and telemetry for production AI systems.
• Contribute to internal standards and best practices for adopting AI tooling, and share knowledge with colleagues.

Required profile

• 5+ years of experience in software engineering, infrastructure, platform or DevOps/SRE roles.
• Experience integrating third-party APIs into production systems (LLM APIs a strong plus).
• Experience with cloud infrastructure (AWS, GCP or Azure) and infra-as-code.
• Hands-on experience with containers and web services (APIs, web apps, backend services).
• Comfortable in both Linux and Windows server environments.
• Pragmatic about buying versus building, with a bias toward iteration over perfection.

Required skills

• Strong scripting and programming ability (Python, Java, etc.).
• Experience with prompt management, retrieval-augmented generation, agentic tooling, MCP or similar.
• Nice to have: LLM observability and eval tooling, self-hosted models, automation or network infrastructure background.

The role is not a research or data science position: no model training, fine-tuning or novel architecture work. Benefits include health and wellbeing support, professional development programs and an inclusive workplace.