How to apply for Senior Backend Engineer (AI Platform)

About AZX, PBC.

AZX is a public benefit corporation founded in 2024 that focuses on AI transformation for critical industries like clean energy, decarbonization, and climate risk. It works with clients such as CBRE, LevelTen Energy, and Flexe, and has been profitable from inception. The company specializes in physics-informed ML and enterprise AI solutions.

About the role

This senior backend engineer role centers on building and owning an LLM gateway that provides a single governed front door over many models, both hosted and self-hosted. You will handle routing, metering, budgets, guardrail composition, and provider adapters. The platform's first users are AZX's own client-facing engineers, so requirements are concrete and feedback is immediate.

A typical day

You might start by reviewing requirements from client-facing engineers, then work on extending the LLM gateway's routing or metering logic. Later, you could test provider adapters for a new self-hosted model or refine guardrail composition. You may also join a meeting to discuss feedback on the platform's performance or cost tracking.

Who AZX, PBC. is looking for

• Has strong backend engineering experience, including designing and operating services that handle routing, metering, and budgets.
• Is familiar with LLM infrastructure concepts such as model routing, guardrails, and provider adapters across hosted and self-hosted models.
• Can work closely with internal users (client-facing engineers) to gather requirements and iterate quickly based on same-day feedback.
• Is motivated by applying AI to climate and sustainability challenges in industries like energy, real estate, logistics, and utilities.

Tips for this application

• Highlight any experience you have with LLM gateways, model routing, metering, or budget enforcement in backend systems.
• Mention specific examples of building services that support internal engineering teams or client-facing engineers.
• Emphasize familiarity with both hosted and self-hosted model deployments, including provider adapters.
• Reference AZX's focus areas (clean energy, decarbonization, climate risk, energy systems) and explain why you want to work on AI infrastructure for those domains.
• Note that AZX is a public benefit corporation and profitable from inception; show that you understand the company's dual focus on impact and financial sustainability.

What to cover in your cover letter

['Your experience owning or contributing to an LLM gateway or similar backend infrastructure with routing, metering, and guardrails.', 'Concrete examples of building services for internal users, especially engineers who deliver client outcomes.', 'Your interest in applying backend engineering to climate and sustainability challenges through AI transformation.', 'How you handle provider/backend adapters across hosted and self-hosted models, and your approach to governed model access.']

Research before applying

• Read about AZX's clients (CBRE, LevelTen Energy, Flexe) and the industries they operate in.
• Understand what physics-informed ML means and how it might relate to LLM infrastructure.
• Look into the challenges of governing LLM usage in enterprises, such as cost tracking, access control, and compliance.
• Check AZX's public benefit corporation status and its implications for company goals and culture.

Likely interview topics

Based on the job description, expect questions about:

• How you would design a routing layer that selects among multiple LLM providers based on cost, latency, or policy.
• Your approach to metering model usage per user, team, or client, and enforcing budgets.
• How to compose guardrails from multiple sources and apply them consistently across models.
• Experience with provider adapters for both hosted APIs and self-hosted models.
• How you would gather requirements from client-facing engineers and iterate quickly based on same-day feedback.

Common mistakes to avoid

• Applying with a generic backend resume that does not mention LLM infrastructure, model routing, or metering.
• Failing to show interest in AZX's focus areas like clean energy, decarbonization, or climate risk.
• Ignoring the internal-user aspect of the role; the platform's first customers are AZX's own client-facing engineers.