Scale AI
January 2024 - April 2026
Jan 2024 - Apr 2026 · 2 yrs 4 mos
Technologies: Node.js, JavaScript, TypeScript, Python, React, Express, PostgreSQL, Redis, RabbitMQ, BullMQ, Prisma, NoSQL, MongoDB, Docker, AI integrations, Vitest, Git, GitHub, GitHub Actions, CI/CD, n8n, Event-driven Architecture, ERP, CRM, APIs, DevOps.
• Engineered a remote Learning Management System (Node.js, React.js, PostgreSQL) with a multi-gateway payment router that retried declined transactions on alternate providers, letting a college with local-only operations expand nationwide while cutting payment failures and lifting revenue.
• Built a WhatsApp chatbot product automating customer service via configurable conversation flows and auto-generated CRM cards per interaction, freeing sales teams to focus on closing and letting marketing scale WhatsApp traffic.
• Built and licensed a multi-tenant AI-powered SDR agent (LangChain, Python, Node.js, PostgreSQL, RAG/pgvector) that qualified and managed leads inside client CRMs via WhatsApp with session-aware AI-to-human handoff, increasing sales and response speed.
• Built an MVP AI agent (LangChain, n8n) for Mercado Livre sellers that auto-answered buyer questions via a RAG knowledge base (PostgreSQL/pgvector) against product listings, responding to buyers in under 5 minutes.
• Designed a multi-tenant platform (React.js, Node.js, RabbitMQ, event-driven architecture) automating hiring and operations for cleaning franchises, with real-time service tracking entirely through WhatsApp and no dedicated app.
• Reduced ERP implementation costs by replacing commercial systems with customized open-source alternatives, saving client companies an average of R$100,000 per implementation.
• Wrote unit tests with Vitest and implemented CI/CD pipelines via GitHub Actions, cutting deployment time from 1 hour to 20 minutes and strengthening regression coverage.
• Owned architecture design across all projects and served as sole DevOps engineer, managing infrastructure on Digital Ocean with Docker.