AI Systems Engineer

About Illumbria

Illumbria exists to discover what AI makes possible and build what comes next. We are drawn to ideas that once seemed out of reach, and to possibilities that people have not yet imagined. Through product design, applied AI, and engineering, we create AI-native applications that expand how organizations work and what they can offer their customers.

We are building a company where curious, ambitious people can help shape both the ideas and their implementation. Our work crosses industries and disciplines, often taking us into problems with no established approach. That calls for independent thinking, experimentation, and the judgment to turn new possibilities into software people can depend on.

We care deeply about what we build and how well we build it. We want people who question assumptions, contribute their own ideas, and take responsibility for seeing them through. The opportunity is to help define what these applications can become—and do the work that brings them into existence.

The Role

We are looking for an AI Systems Engineer to design and build applications with AI at their core. You will help turn ambitious ideas into working software, making decisions about architecture, developing the systems behind the application, and refining how they perform in real use.

The work spans AI agents, data, integrations, and the application itself. You will explore approaches where established patterns may not exist, evaluate what works, and carry your decisions through implementation. This is a hands-on engineering position with meaningful ownership of both how the systems are built and how well they work.

Responsibilities

• Work with clients to understand their needs, assess technical feasibility, and explain proposed approaches and tradeoffs.
• Evaluate emerging models and techniques through focused experiments, and apply them where they improve the systems we build.
• Build the application around AI workflows, including user interfaces, APIs, data models, integrations, and background processing.
• Build AI agents and workflows that use tools, manage context, verify results, and recover from interruptions or failures.
• Develop data-processing and retrieval systems that give applications accurate, relevant information with traceable sources.
• Develop reusable components and internal tooling that improve how we build and maintain applications.
• Build evaluations and regression tests that measure changes to models, prompts, and application behavior before they reach production.
• Improve response times and operating costs while maintaining quality, and make system performance measurable.
• Review and test code, including work produced by AI coding agents, and document decisions so others can maintain and extend the systems.

How We Work

AI coding agents are part of our everyday engineering workflow. You will direct their work, evaluate what they produce, and take responsibility for the software you ship. We value knowing when to delegate, when to investigate further, and when to work through a problem yourself.

You will have clear priorities and the autonomy to decide how to approach them. Bring ideas, question assumptions, and communicate the reasoning behind your decisions.

Requirements

• A record of shipping production software you can point to and explain in detail.
• Experience building applications on top of modern LLM APIs: structured outputs, tool calling, multi-step agent runs and the failure modes that come with them.
• Strong backend engineering: data modeling, APIs, async and long-running work, and integration with external services.
• Ability to learn unfamiliar systems and work through technical problems independently.
• Judgment about when AI is the right tool, when deterministic code is, and how to put guardrails around the boundary.
• Experience using AI coding agents, or the ability and willingness to adopt them while independently verifying their output.
• Clear written communication.

Helpful Experience

Useful, but not required:

• TypeScript and a modern web framework, or an equivalent stack you could carry over.
• Word and PDF internals (OOXML, PDF text extraction, OCR).
• AI evaluation: benchmark datasets, regression testing, human-in-the-loop review.
• Financial services, legal or other compliance-sensitive software, including SOC 2-aligned practices.
• Cloud deployment and operations on AWS, Cloudflare or similar.

What We Offer

• Flexible hours and competitive pay.
• Responsibility for technical decisions, with room to propose approaches and carry them through implementation.
• Work across different applications and industries as projects develop.