Engenheiro(a) de Prompts / Desenvolvedor(a) de Agentes de IA
J
JobgetherAI software
Based in BrazilFull-Time
Salary not disclosed
Apply NowOpens the employer's application page
Job Details
- Languages
- Excellent Portuguese writing skills
- Required Skills
- PythonGitCI/CDPrompt Engineering
Requirements
- Have hands-on prompt engineering experience in real-world projects using multiple LLM families, such as OpenAI/GPT, Anthropic/Claude, Google/Gemini, and Meta/Llama.
- Have practical experience developing AI agents within squads, including orchestration, tool integration, and context-memory management.
- Have experience defining metrics, test cases, and evaluation approaches for AI response quality and reliability.
- Demonstrate excellent Portuguese writing, including clarity, grammar, precision, tone, and concise instruction-writing.
- Be familiar with Git and existing CI/CD pipelines for version control and deployment.
- Be adaptable and willing to learn and operate new proprietary AI platforms and technologies.
- Be able to critically evaluate AI-generated content and iterate on prompts and workflows based on test results.
- Bring collaboration, analytical thinking, attention to detail, and curiosity about emerging AI technologies.
- Experience with RAG, particularly evaluating responses against documentary sources, is a plus.
- Knowledge of Python for automating LLM testing and evaluation is a plus.
- Experience handling sensitive data under LGPD requirements or working in healthcare environments is a plus.
Responsibilities
- Design and configure system prompts, instructions, and conversational workflows for AI agents across use cases and LLM capabilities.
- Evaluate and benchmark multiple LLM families, balancing response quality, cost, latency, and model-specific characteristics.
- Build and refine AI agents within collaborative squads, including orchestration, tool usage, and context-memory management.
- Define response-quality criteria and conduct iterative testing of AI-generated outputs.
- Implement guardrails, out-of-scope response handling, and mitigations for risks such as prompt injection.
- Document prompt engineering standards, versioning practices, and reusable patterns.
- Collaborate with technical teams to publish and maintain solutions through existing Git and CI/CD workflows.
- Investigate new AI platforms and techniques and adapt to proprietary tools and evolving project requirements.
View Full Description & ApplyYou'll be redirected to the employer's site
