We're looking for someone who builds and ships internal, production-grade web applications end-to-end — from a business problem to a live URL people use daily — by directing AI coding agents (Claude Code, GitHub Copilot) rather than hand-writing every line.
This isn't "prompt and hope": it's owning the full lifecycle — architecture calls, integration debugging, production incidents, and the judgment to know when the AI's output needs to be checked rather than trusted.
You'll work largely solo, moving fast between ideas, working prototype, and production deployment, while integrating with real enterprise systems (SSO, mailboxes, databases, cloud infrastructure) that don't forgive sloppy assumptions.
How You’ll Make an Impact (key responsibilities of role)
• Take a business need directly from a stakeholder (a team lead, an exec, a support team) and turn it into a working Next.js/React app without a formal spec or a dedicated developer team.
• Direct AI coding agents to build features, then verify their output against how the system actually behaves — not just whether it compiles.
• Integrate with enterprise identity and data systems: SSO (Cognito + Entra), Microsoft Graph (mailbox/calendar access), shared internal database APIs, cloud AI services (AWS Bedrock)
• Own deployment: CI/CD pipelines, containerized deploys, environment configuration — enough infra literacy to read a failed pipeline and fix it yourself
• Debug production issues by building your own diagnostics (an admin page, a probe endpoint) rather than waiting on someone else's logs
• Recognize the boundary between "I can fix this myself" and "this needs IT/platform team” and write a clear, specific ask when it's the later, instead of thrashing on a workaround
• Test in production and iterate from real usage rather than over-planning up front
• Keep documentation current enough that the next person (or the next AI session) doesn't have to rediscover what you already figured out
What You Bring (required qualification and skill sets)
Technical
• Bachelor’s or master’s degree in computer science, Software Engineering, Robotics, or a related field with minimum 5-6 years of relative experience
• Full-stack web development: React/Next.js, TypeScript, REST APIs
• Cloud fundamentals: AWS IAM, ECS/containers, CloudFormation basics — enough to reason about who owns what permission
• Enterprise auth: OAuth/OIDC, SSO integration (Cognito, Entra ID)
• SQL / relational data modeling, working within a shared/managed database rather than standing up your own
• Git/GitLab, CI/CD pipeline literacy
AI-assisted development ("vibe coding") specifically
• Fluency directing AI coding agents to build real, deployed software — not just toy scripts
• The judgment to catch when an AI's output is subtly wrong (e.g., an IAM policy written for the wrong model version) rather than shipping it as-is
• Comfortable debugging by reading actual system behavior (live API calls, logs, diagnostics) instead of guessing
• Knows the difference between a code problem and an infrastructure/permissions problem — and doesn't burn cycles trying to code around the latter
Working style
• Direct communicator — wants the one-sentence answer, not a hedge
• Ships fast, tests in prod, iterates from what breaks
• Comfortable being both the builder and the product owner for what you ship
• Persistent through ambiguous, multi-day infrastructure blockers without losing momentum on everything else
Nice to have
• Design-system awareness (brand consistency, accessibility)
• Experience building analytics/reporting features (charting, data taxonomies, dedup pipelines)
• Comfort coordinating directly with IT/platform teams on access and permissions requests