Note: The job is a remote job and is open to candidates in USA. Revley is building an operating system for online stores that integrates commerce, payments, subscriptions, analytics, and business operations. The AI Application Engineer will build production AI features at the application layer, including agent orchestration, retrieval, tool calling, and workflow automation. The role also involves developing secure full-stack systems, integrations, observability, and reliable AI-triggered product experiences.
Responsibilities
- Build production AI features that understand user intent, retrieve the right business context, use tools, and complete multi-step workflows
- Design agent orchestration, tool and function calling, structured outputs, context management, memory, retrieval, and human-in-the-loop approval patterns
- Connect AI applications to Revley OS through secure APIs, MCP tools, internal services, and event-driven workflows
- Build backend services and customer-facing product surfaces across TypeScript, NestJS, Next.js, React, PostgreSQL, and AWS
- Design reliable integrations using OAuth, webhooks, queues, scheduled jobs, and third-party APIs
- Implement tenant isolation, scoped authorization, idempotency, audit logs, rate limits, and safe execution for AI-triggered actions
- Create evals, tracing, monitoring, and feedback loops for accuracy, tool selection, latency, reliability, and task completion
- Debug the full execution path—from model output and orchestration through APIs, databases, and user interfaces
- Work directly with product and engineering to turn ambiguous operating problems into scoped, production-ready systems
Skills
- Experience shipping AI-powered applications, agents, copilots, or workflow automation into production
- Strong knowledge of the AI application layer: tool/function calling, structured outputs, retrieval, context engineering, orchestration, evals, and observability
- Strong full-stack engineering experience with TypeScript and modern web applications
- Experience building APIs, distributed workflows, data models, OAuth integrations, and webhook systems
- Strong understanding of production concerns including security, permissions, multi-tenancy, testing, failure recovery, and monitoring
- Ability to evaluate AI systems empirically rather than relying only on prompt iteration
- Product judgment, clear communication, and evidence of owning meaningful systems from design through production
- You do not need experience training foundation models
- United States
- MCP or comparable tool-integration protocols
- NestJS, Next.js, React, PostgreSQL, AWS, Vercel, queues, and serverless systems
- Commerce, payments, subscriptions, fintech, or operational software
- Conversational interfaces, workflow engines, or approval-based automation
- Prompt-injection defense and capability-based authorization
Benefits
- Build AI that takes useful, authorized action—not just generates text.
- Work on technically demanding problems across AI, commerce, payments, and business operations.
- Own meaningful product and infrastructure surface area in a small engineering team.
- Ship quickly, observe real usage, and help define the architecture of Revley OS.
Company Overview