Engenheiro(a) de IA - Aplicações Agênticas (Pleno)
New
J
JobgetherGenerative AI
Based in BrazilContractMiddle
Salary not disclosed
Apply NowOpens the employer's application page
Job Details
- Languages
- Professional proficiency in English or Spanish
- Required Skills
- PythonRESTful APIsPrompt EngineeringLangChain
Requirements
- Bachelor’s or postgraduate education in Data Science, Computer Science, Statistics, Engineering, Mathematics, Economics, or a related field.
- Professional proficiency in English or Spanish, with the ability to communicate effectively in an international client environment.
- Strong programming skills in Python and/or C#/.NET.
- Experience with testing, version control, software contracts, and writing maintainable, readable code.
- Hands-on experience developing LLM-powered applications, including tool or function calling, agent orchestration, prompt engineering, and strategies for managing non-deterministic model behavior.
- Understanding of system integration and REST API development or consumption, including OpenAPI/Swagger, OAuth2 authentication, idempotency, webhooks, and callbacks.
- Experience enabling secure LLM access to data, including natural-language-to-query workflows with scope controls, validation, permissions, and auditability.
- Understanding of application security for LLM-based systems, including prompt injection, data leakage, and permission management at the tool level.
- Experience with MCP, LangChain or Semantic Kernel, Azure OpenAI, multi-agent architectures, Human-in-the-Loop patterns, agent observability, Kafka, Azure Service Bus, or .NET is considered a strong advantage.
- Familiarity with regulated environments, LGPD requirements, and auditing of automated decisions is valuable.
Responsibilities
- Develop specialized AI agents coordinated by supervisors, implementing agentic loops such as ReAct and maintaining auditable records of decisions and actions.
- Build MCP Servers that connect AI agents with operational APIs and machine learning models through well-defined OpenAPI contracts.
- Implement MCP Hosts using LangChain or Semantic Kernel and integrate LLM capabilities through Azure OpenAI.
- Design and implement Human-in-the-Loop workflows for decisions requiring human validation, oversight, or intervention.
- Establish secure agentic architectures with agent-specific identities, OAuth2 authentication, tool-level permissions, idempotency, prompt-injection protection, and safeguards against PII leakage.
- Apply PromptOps practices by versioning prompts and implementing agent observability through LangSmith or similar platforms.
- Integrate AI applications with event-driven architectures using technologies such as Azure Service Bus and Kafka.
View Full Description & ApplyYou'll be redirected to the employer's site
