AI-First Backend Engineer (Brazil)

AI-First Backend Engineer Claude, OpenAI, Agents, Loops and Production AI LOCATION: Brazil TYPE: Full-time TEAM: WowLoop Build AI that makes real commercial decisions, not another chatbot. WowLoop is an AI-first membership and retention platform. Its intelligence layer, Mercer, helps merchants design, evaluate, approve, launch and improve membership programs using real pricing, margin, inventory, customer and operational data. Mercer must be able to say launch it, fix it or stop , and explain why. It must distinguish drafts from approved facts, revenue from profit, and operational possibility from commercial value. We need an AI-first backend engineer who has already built real systems with Claude, ChatGPT/OpenAI or comparable models . This is not a role for someone whose AI experience is limited to prompting a chatbot or generating code. What You Will Own Build Mercer’s AI reasoning and decision systems. Create multi-step agent, tool-calling and workflow loops. Connect LLMs to APIs, databases, business rules and live product data. Build evaluation, feedback, correction and human-approval loops. Design structured outputs reliable enough for production workflows. Manage prompts, context, model selection, retries, cost, latency and failure handling. Build the backend services and APIs that power WowLoop. Help define the AI architecture, not simply implement tickets. You Must Have Shipped at least one real AI product, agent or workflow used by actual users. Hands-on experience with the Claude API, OpenAI API or another production LLM platform. Built at least one meaningful loop, such as an agent and tool-use loop, evaluation loop, review and correction loop, human-in-the-loop approval process or outcome-feedback loop. Experience testing AI quality rather than relying on subjective prompt results. Strong backend engineering, API design, data modelling and debugging skills. Daily experience with Cursor, Claude Code, ChatGPT, Copilot or similar AI development tools. Professional English: CEFR B2+, IELTS 6.5+ or equivalent. Our Environment Azure, .NET microservices, Redis, SQL Server, MongoDB and React. C# experience is valuable, but demonstrated AI engineering and strong backend judgment matter more than years in one language. You must be willing and able to work effectively in a Microsoft-based production stack. Apply With Evidence One AI product, agent or loop you built. A demo, GitHub repository or architecture overview. How the system used models, prompts, tools, data and evaluations. What failed, how you measured it and how you improved it. The business or user result. We hire builders. Show us working AI, not AI vocabulary.