Sr. Agentic AI Software Engineer

country: US all locations: [Farmington Hills or Remote (US only)] commitment: Full-time department: location: Farmington Hills or Remote (US only) team: Engineering Required Skills:: Strong experience in Ruby, with working knowledge of Python for machine learning and statistical analysis, and familiarity with Go for high-performance needs Demonstrated daily use of Claude Code as a primary agentic coding tool, with production PRs to show for it, not just experimentation Experience building and coordinating multi-agent workflows with proper checkpointing, reproducibility, and cost control Ability to author and maintain context documents (CLAUDE.md, architecture maps, runbooks) that make a codebase AI-navigable Experience designing evaluation harnesses to catch regressions and measure AI-assisted velocity Disciplined prompt engineering: you design prompts with intent, iterate systematically, and document what works Experience with RAG systems, MCP servers, or multi-agent orchestration Security-first mindset around AI workflows: data classification, secrets hygiene, prompt injection awareness Strong fundamentals in system design, debugging, and architecture reasoning Ability to teach and share knowledge in ways that genuinely move a team forward Excellent communication and collaboration skills Job Responsibilities:: Demonstrate total fluency in Ruby on Rails and at least one modern front-end environment, while maintaining the confidence and adaptability to work in any codebase on any platform by leveraging AI coding agents effectively. Operate daily with Claude Code using structured, repeatable workflows: plan, decompose, implement, test, ship Know when to trust AI output and when to verify it. You are the quality gate, not the rubber stamp Review AI-generated code with discipline: know what patterns to trust, where to focus attention, and when to regenerate vs. fix manually Use AI for planning before code exists: technical design, task decomposition, risk identification, and effort estimation Serve as a technical resource for AI-augmented workflows across the engineering team, sharing playbooks, patterns, and tooling that help other developers level up Build reusable agent skills, hooks, and guardrails that raise the team-wide baseline Collaborate with product managers, front-end developers, and linguists across cross-functional teams Participate in code reviews, sprint planning, and agile processes Continuously research and evaluate new AI tools, models, and techniques with a framework for deciding what to adopt Who You Are:: You are a top-tier engineer who can perform the work of your agents and more; you use AI not to compensate for a lack of skill, but to amplify your already superior capability and accelerate delivery. You do not treat AI as a shortcut. You treat it as infrastructure. You have built workflows, not just prompts. You understand context engineering. You know when to let the agent run and when to take the wheel. You verify the output because you own the result. You measure your process and improve it. When AI produces poor results, your first question is what you can change, not what AI cannot do. You are also someone who takes initiative and ownership. At Mango, we hold ourselves accountable to results, we find a way, and we passionately pursue excellence and continuous improvement. If that describes how you already work, you will feel at home here.