Staff Software Engineer (AI Native Development)

Payroll Integrations is hiring a Staff Software Engineer to lead its shift to AI-native software development. The role defines engineering standards and governance for AI-assisted delivery, builds the supporting toolchain, and ships production software while ensuring correctness, security, and team adoption.

Responsibilities:

• Help set the standard. Work with engineering leadership and our AI advisor to define how Payroll Integrations builds software with AI: the workflow, the guardrails, the review bar, and what done means for agent-produced code
• Build the toolchain. Design and maintain the skills, agents, prompts, evals, and CI checks that make agentic delivery repeatable and trustworthy across teams
• Prove it in production. Ship features across our teams using the workflow you define, and use what you learn to improve it
• Own correctness. Hold every change, yours or an agent’s, to clean architecture boundaries and to tests that cover the real risk: money, deductions, permissions, and integration boundaries
• Measure what works. Instrument AI-assisted delivery, from cycle time and defect rates to review findings and cost, and report plainly on what the tools are and are not doing
• Govern responsibly. Help shape policy for what data may reach which models, how AI-generated code is reviewed and attributed, and how we stay inside our security and compliance obligations
• Lead through others. Coach engineers and product managers on effective, honest use of AI, and raise the whole team’s judgment, not just its throughput
• Decide at your level. Make and document the architecture and tooling decisions that are expensive to reverse, and push back with evidence when a request does not hold together

Requirements:

• Ten or more years building production software, including time operating at staff or principal level, with deep TypeScript or Node.js experience on a cloud platform
• You understand how large language models work well enough to reason about them: context engineering, tool use and agent loops, caching, evals, and the ways models fail
• You have built with them, not just used them
• You have run an agentic workflow in production: specs that agents build from, tests they must pass, evals that catch regressions, and review that catches what looks right but is not
• You can show the results
• You have set engineering standards that a team actually adopted, and you know the difference between a rule people follow and a document people ignore
• You practice test-driven development and treat coverage as a map of risk rather than a number to hit
• You can explain Clean Architecture or a close relative, you have designed serverless cloud systems, and you have kept business logic out of handlers and repositories under deadline pressure
• You know the risks AI tooling and AI-generated code introduce, from prompt injection to leaked secrets and hallucinated dependencies, and how to guard against them
• You write clear specs and decision records, and you can explain a technical trade-off to leadership, a product manager, or a client
• Experience in fintech, payroll, HR tech, benefits administration, or retirement
• Experience building MCP servers, agent skills, or eval harnesses that other engineers depend on
• Experience integrating with third-party APIs, SFTP file exchange, or payroll and HRIS platforms
• Angular and front-end experience
• Writing, talks, or open-source work on AI-assisted engineering