How to apply for Infrastructure Engineer

About Elicit

Elicit is a public benefit company building an AI research assistant that helps researchers make better decisions. The product is a scalable ML system that prioritizes systematicity and transparency, with supervision of process rather than outcomes. This is a chance to work on infrastructure for high-stakes scientific decision-making.

About the role

You will design and operate cloud infrastructure for an AI research platform used in scientific decision-making. You will own AWS and GCP infrastructure, including Kubernetes, databases, CI/CD, and single-tenant deployments for enterprise customers. You will also build observability, incident response, and disaster recovery practices, and drive SOC 2, NIST AI framework, and EU Cyber Resilience compliance.

A typical day

A typical day might involve reviewing Terraform changes, troubleshooting Kubernetes issues across AWS and GCP, and improving observability dashboards. You could also work on single-tenant deployments for enterprise customers, respond to incidents, or advance compliance documentation for SOC 2 or NIST AI. Since the team is remote, you would coordinate with engineers and researchers asynchronously.

Who Elicit is looking for

• 5+ years of infrastructure, SRE, or platform engineering experience.
• Strong hands-on expertise with Terraform and Kubernetes in production.
• Experience managing multi-cloud environments (AWS and GCP) and single-tenant enterprise deployments.
• Enthusiastic practitioner of agentic AI tools like Claude Code, Cursor, or Devin to multiply infrastructure impact.
• Familiarity with compliance frameworks such as SOC 2, NIST AI, or EU Cyber Resilience, and building observability, incident response, and disaster recovery practices.

Tips for this application

• Highlight specific examples of owning AWS and GCP infrastructure, including Kubernetes and Terraform, in production.
• Describe any experience with single-tenant deployments for enterprise customers, as this is a key responsibility.
• Show how you have used agentic AI tools (Claude Code, Cursor, Devin) in your infrastructure work, with concrete outcomes.
• Mention any direct work on SOC 2, NIST AI framework, or EU Cyber Resilience compliance implementation.
• Explain your approach to building observability, incident response, and disaster recovery with proactive monitoring and SLA tracking.

What to cover in your cover letter

['Your experience designing and operating multi-cloud infrastructure (AWS and GCP) with Kubernetes and Terraform.', 'How you have built observability, incident response, and disaster recovery practices with proactive monitoring and SLA tracking.', 'Your hands-on work with compliance frameworks like SOC 2, NIST AI, or EU Cyber Resilience.', 'Concrete examples of using agentic AI tools to improve infrastructure work.']

Research before applying

• Read the Elicit website and any public documentation to understand the AI research assistant product and its users.
• Look for Elicit engineering blog posts or talks about their infrastructure, ML systems, and approach to transparency.
• Research the NIST AI framework and EU Cyber Resilience requirements to speak to compliance implementation.
• Check Elicit's public benefit company structure and how it influences their product and infrastructure decisions.

Likely interview topics

Based on the job description, expect questions about:

• How you design and manage Kubernetes clusters across AWS and GCP for high-stakes applications.
• Your approach to Terraform module design and state management in multi-cloud environments.
• How you build observability, incident response, and disaster recovery with SLA tracking.
• Your experience implementing SOC 2, NIST AI framework, or EU Cyber Resilience compliance.
• How you use agentic AI tools like Claude Code, Cursor, or Devin in your daily infrastructure work.

Common mistakes to avoid

• Listing generic infrastructure experience without tying it to AWS, GCP, Kubernetes, and Terraform in production.
• Ignoring the compliance requirements (SOC 2, NIST AI, EU Cyber Resilience) or treating them as someone else's job.
• Claiming enthusiasm for agentic AI tools without giving specific examples of how you used them to improve infrastructure work.