Staff Software Engineer (AI Native Development)

Payroll Integrations is a technology platform that powers secure data connections between payroll providers and financial and benefits platforms. The Staff Software Engineer will lead AI-native development by establishing engineering standards, building agentic delivery toolchains, shipping production software, and coaching teams on effective AI use. The role also emphasizes correctness, security, architecture, testing, and responsible AI governance.

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.Standards
• Build the toolchain. Design and maintain the skills, agents, prompts, evals, and CI checks that make agentic delivery repeatable and trustworthy across teams.Toolchain
• Prove it in production. Ship features across our teams using the workflow you define, and use what you learn to improve it.Delivery
• 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.Quality
• 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.Standards
• 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.Standards
• 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.Leadership
• 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.Leadership

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