______________________________________________________________________

Role: AI & Workflow Automation Engineer (Claude / Anthropic Ecosystem)

Priority Location: Brazil, Mexico, Colombia, or South Africa (LATAM & South Africa focus)

Architect, test, and deploy custom Claude-based AI agents and agentic workflows utilizing Anthropic’s API, Tool Use (function calling), and Prompt Caching features.
Design and implement robust integrations between LLM APIs and core business systems (e.g., CRMs, ERPs, databases, webhooks, or third-party APIs) using Python, Node.js, or middleware tools (e.g., Make, n8n, LangChain, LlamaIndex).
Craft, benchmark, and continuously optimize complex system prompts and structured outputs (JSON/Markdown) to maintain low error rates and consistent execution.
Build automated testing pipelines to evaluate agent accuracy, latency, guardrails, and failure recovery across varied operational scenarios.
Collaborate closely with product managers and non-technical stakeholders to identify workflow bottlenecks and translate manual procedures into reliable, automated AI workflows.
Implement error handling, fallback logic, rate-limit management, and telemetry logging to ensure enterprise-grade stability and security across all deployed agents.
Maintain clear technical documentation for agent topologies, prompt schemas, and API integration endpoints.

Minimum 2 years of hands-on experience building production-ready applications or automated workflows using Large Language Models (LLMs), with specific expertise in the Anthropic/Claude platform (Required).
Strong programming background in Python or TypeScript/JavaScript for API integration, asynchronous execution, and data parsing (Required).
Deep understanding of LLM primitives: context windows, function calling / tool use, system prompts, temperature controls, and structured JSON outputs (Required).
Proven experience integrating REST APIs, webhooks, and database endpoints to enable AI agents to execute multi-step actions (Required).
Professional fluency in written and spoken English for technical collaboration and documentation (Required).
Experience with orchestration frameworks (e.g., LangChain, AutoGen, CrewAI, or LlamaIndex) or visual automation builders (n8n, Make) (Preferred).
Familiarity with Retrieval-Augmented Generation (RAG) architecture, vector databases (e.g., Pinecone, Qdrant, Pgvector), and token usage optimization (Preferred).
Previous experience working as a remote independent contractor within distributed, multi-timezone environments (Preferred).