Engenheiro de IA, Pleno

This is a fully remote position, open to applicants in Brazil.

📋 Description

• Design, develop, and enhance **prompts, chains, and conversational flows** for LLM-based applications, ensuring consistency, accuracy, and alignment with business context;

• Define and implement **prompting strategies** (zero-shot, few-shot, chain-of-thought, tool use), assessing impacts on quality, cost, and latency;

• Create and maintain **quality evaluation mechanisms**, including automated metrics, A/B testing, and qualitative analysis, utilizing validation datasets and benchmarks;

• Establish and enforce **guardrails**, response standards, and security policies aimed at mitigating hallucinations, biases, inappropriate responses, and off-scope behavior;

• Collaborate in defining **security and privacy strategies**, preventing the leakage of sensitive data (PII), and ensuring compliance with internal policies and regulations;

• Engage in the **continuous improvement of AI applications**, analyzing logs, user feedback, and evaluation results to refine models and flows;

• Work collaboratively with **product, engineering, and UX teams**, supporting the integration of AI solutions into digital products, APIs, and chat experiences;

• Document technical decisions, usage standards, and best practices, contributing to the scalability and maintenance of the developed solutions.

⛳️ Requirements

• Practical experience with LLMs (specifying models, use cases, and results);

• Prompting techniques, structuring, few-shot, tool use;

• Evaluation: quality metrics, A/B testing, validation datasets;

• Security and privacy: prevention of data leakage, PII, policies;

• RAG (embeddings, chunking, retrievers), vectors, and relevance;

• Integration with products (APIs, chat UX, logging).

🏝️ Benefits

• Position also available for people with disabilities (PcD).