Staff Engineer – AI-Native Commerce

Phoenix is an autonomous commerce platform for direct-to-consumer founders that unifies storefronts, checkout, payments, and operations with AI agents. The Staff Engineer will put AI at the core of Phoenix's storefront platform by building and evaluating AI agents, designing autonomous product systems, making architectural decisions, writing production code, and improving AI-native engineering practices and technical quality.

Responsibilities:

• Take Phoenix's copy-site and AI editing capabilities from impressive to trusted
• Design the agent's tools and execution loop, determine where merchants should confirm actions before execution, and build the evals, tracing, observability, and cost tracking required to prove that changes improve the system
• Harden and expand Phoenix's MCP infrastructure and create new tool surfaces that allow agents to reliably build and modify pages, sections, components, styles, and data bindings
• Build evaluation systems against real commerce experiences so regressions are identified before merchants encounter them
• Build the infrastructure that allows AI agents to generate storefront experiences, validate them against quality gates, deploy them, measure their performance, and continuously optimize them using conversion data
• Design autonomous workflows with the appropriate contracts, guardrails, verification, and human checkpoints so agents can operate reliably in production
• Design the APIs, data models, service boundaries, and technical architecture behind Phoenix's AI-native storefront platform
• Work alongside the Engineering Lead to make pragmatic architecture decisions across TypeScript/Node.js services, React applications, Postgres, MongoDB, Redis, and related systems
• Balance rapid execution with architectural decisions that will continue to work as Phoenix scales
• You will work this way from day one: writing specifications agents can execute, running agents in parallel, reviewing and verifying their output, and continuously improving the development system itself
• Build leverage that extends beyond your own output through custom agents, skills, hooks, MCP servers, automated verification, and AI-assisted development infrastructure
• Raise the verification bar for code written by both engineers and agents
• Build deterministic fixtures, Playwright E2E testing, golden-file rendering tests, integration tests against real services, visual regression systems, and automated quality gates
• Own observability, tracing, SLOs, and operational reliability for the systems you build
• Author ADRs, RFCs, and technical specifications and lead design and code reviews that improve engineering quality across the organization