Full-Stack Engineer — AI Agents & Automation

EHMA builds the software behind Everpower’s family of products. We’re looking for a full-stack engineer with a real knack for AI — someone who writes solid, production software and enjoys putting large language models and agentic systems to work on real problems. This is not a research role: we’re not looking for an AI PhD, we’re looking for a strong, versatile developer who is hands‑on with LLMs and wants to own the AI‑powered features and automations that are becoming core to how our business runs.

You’ll work across the stack and across products, closely with our Director of Engineering.

• Build and maintain AI-powered features and agentic automations across our products and internal operations.
• Take day‑to‑day ownership of one or more of our AI systems — improving quality, expanding capabilities, and keeping them reliable.
• Work across the stack — backend services and the interfaces and integrations around them — to ship complete features, not just prototypes.
• Build the evaluation and testing that keep AI systems trustworthy: measuring quality, catching regressions, and enforcing safe, accurate output.
• Integrate LLMs with our data and tools through retrieval, tool/function calling, and well-structured pipelines.
• Help make our AI tooling documented, tested, and maintainable — so critical systems don’t depend on any one person.

• Owning the quality and release cadence of a live AI system.
• Shipping a new AI-powered feature or automation end to end.
• Strengthening how we evaluate and safeguard AI output.

• 2–4 years of professional software engineering, comfortable across the stack (strong backend, capable on the frontend).
• Hands‑on experience building with large language models — you’ve shipped something real with them, not just experimented.
• Solid engineering fundamentals: you write tested, maintainable code and think about correctness and edge cases.
• Curiosity and product sense for where AI genuinely helps — and where it doesn’t.
• Clear communication and comfort working closely with a technical lead.

• Experience with agent frameworks, tool/function calling, or multi‑step agentic workflows.
• Retrieval‑augmented generation (RAG), embeddings, and vector/keyword search.
• C#/.NET (our primary backend) and/or a major LLM platform (Azure OpenAI or equivalent).

EHMA — the technology arm of Everpower

Director of Engineering

Tempe, AZ - Onsite (No Remote, No Hybrid, No WFH)

Full‑time