Staff AI Engineer

SpotOn is a company focused on empowering independent restaurants through innovative technology solutions. They are looking for a Staff AI Engineer to lead the architecture and implementation of AI-assisted engineering practices within their fintech organization, driving the development of automated systems and workflows.

Responsibilities:

• Be the AI engineering technical authority
• Drive The Architecture And Implementation Of Three Concurrent Projects
• Automated PR Review - A risk-tiered AI review pipeline where AI handles first-pass review on every PR, with an escalation framework you design that determines when human review is mandatory (payment logic, PCI-scoped code, compliance paths) versus when AI approval is sufficient
• Define the criteria
• Tune the thresholds
• Prove it works
• Android and API Test Automation - A fully automated test harness for our Android-native POS and KDS products: virtual device provisioning, test execution, teardown, AI-assisted failure analysis - replacing multi-day manual QA cycles with sub-30-minute automated runs
• AI Agent Swarms - A multi-agent pipeline that takes a PRD and produces a tested, reviewed pull request for tier 1-3 engineering work (nits, bugs, small features). You design the agent orchestration - planner, coder, tester, reviewer - and the blast-radius controls that keep it safe in fintech
• Evangelize by shipping, not by presenting
• Demo to engineering leads every two weeks with real metrics
• Turn skeptics into adopters by showing them their own time savings
• Define the metrics framework: developer hours recovered, cycle time compression, quality ratios, cost per deployment
• Report to the CTO monthly with quantified value delivered
• If something isn't working, the data tells you before anyone else

Requirements:

• 5+ years building developer tools, infrastructure, or platform engineering systems - you understand how engineering orgs actually work at scale, not just how AI demos work
• 2+ years working deeply with LLMs and AI agents in production - not side projects, not hackathons. Production systems handling real code, real reviews, or real task automation. You can speak to failure modes, cost optimization, prompt engineering patterns, and model selection trade-offs from experience
• Designed and operated AI-assisted code generation or review systems serving 50+ engineers or 30+ repositories. You know the difference between 'cool demo' and 'trusted by the team.'
• Built multi-agent or agentic workflows - task decomposition, agent specialization, orchestration, error recovery. You've worked with Claude Code, OpenAI Assistants, LangGraph, CrewAI, AutoGen, or custom agent frameworks at a level where you can articulate why you chose one approach over another
• Hands-on with AI coding tools daily - Claude Code, Cursor, GitHub Copilot, Windsurf, Codeium - you don't just use them, you push their limits. You know where they excel and where they hallucinate. You have a personal workflow that makes you measurably faster and you can teach it
• Required Domain Expertise (at Least Two Of Three): Fintech / payments / regulated industry - you understand what PCI-DSS, SOX, or HIPAA compliance means for code review and deployment. You know why 'just let the AI merge it' isn't an option for payment logic
• Mobile / Android testing automation - Espresso, UI Automator, or AI-native tools (Drizz, mabl, testRigor). Experience with emulator farms, device provisioning, and CI-integrated test pipelines
• Multi-agent system architecture - you've built systems where multiple AI agents collaborate on decomposed tasks with coordination protocols, fallback strategies, and human-in-the-loop checkpoints