AI Product Engineer

Role Summary

We are looking for someone to work as an AI Product Engineer, responsible for transforming Artificial Intelligence capabilities into products, agents and workflows capable of solving real business problems.

The position combines software engineering, architecture, applied AI and product vision. We're not looking for someone focused solely on experimenting with prompts or creating proofs of concept. We expect a person capable of identifying opportunities, designing solutions, implementing products and effectively putting them into production.

This person should work mainly on building AI Agents, agentic workflows, execution loops, internal tools and products based on LLMs, integrating AI models into existing systems, APIs, knowledge bases and processes.

We also expect sufficient autonomy to understand and participate in decisions related to the infrastructure and operation of these applications, especially in AWS environments, including serverless applications and Lambda Functions.

Responsibilities

Product development with AI

• Design and develop products and features based on Artificial Intelligence.
• Transform business needs into solutions using LLMs, agents, automations and workflows.
• Develop complete applications, considering frontend when necessary, backend, integrations, persistence, infrastructure and AI layer.
• Quickly create prototypes to validate hypotheses and evolve them into robust solutions when there is evidence of value.
• Work closely with product, engineering and business areas to identify opportunities where AI can generate concrete gains.
• Evaluate when to use generative AI and when a traditional software solution is more appropriate.
• Participate in the entire product cycle, from discovery and experimentation to deployment, observability and evolution in production.

AI Agents

• Design and implement specialized agents capable of performing tasks using context, tools and business rules.
• Define responsibilities, tools, context and limits of agent action.
• Develop multi-agent architectures when there is a real need for specialization or separation of responsibilities.
• Create handoff and coordination mechanisms between agents.
• Implement agents capable of interacting with APIs, databases, internal systems and external tools.
• Develop strategies for managing context, memory and state.
• Define validation and control mechanisms for actions performed by agents.
• Create fallback and recovery mechanisms when an execution does not produce the expected result.

Agentic Workflows and Loops

• Develop workflows in which AI models perform structured sequences of analysis, action, validation, and review.
• Build loops involving steps such as planning, execution, evaluation and correction.
• Implement workflows with multiple steps and conditional decisions.
• Design processes with human-in-the-loop when decisions require human validation.
• Create retry, timeout, fallback and error handling mechanisms.
• Work with synchronous and asynchronous workflows.
• Define clear criteria for starting, continuing and closing loops.
• Avoid undefined behaviors through operational limits and objective completion criteria.
• Evaluate when a deterministic workflow is more appropriate than a fully agentic architecture.

Prompt Engineering and Context Engineering

• Design structured prompts for different types of agents and tasks.
• Work with system prompts, templates, examples and structured instructions.
• Develop context engineering strategies, ensuring that each model receives the information necessary for its task.
• Define structured input and output formats.
• Implement structured outputs and schema validations.
• Create strategies to reduce hallucinations and increase the predictability of responses.
• Manage context windows and context retrieval, summarization and compression strategies.

Tool Calling and integrations

• Create tools that can be used by agents to perform actions on external systems.
• Integrate models with REST APIs, GraphQL, databases, enterprise systems and SaaS platforms.
• Implement Function Calling and equivalent tool execution mechanisms.
• Develop integrations using standards such as Model Context Protocol (MCP) when applicable.
• Define clear contracts between agents and tools.
• Implement authentication, authorization and access controls.
• Ensure agents have only the permissions necessary for their responsibilities.

RAG and knowledge

• Develop solutions using Retrieval-Augmented Generation (RAG).
• Error 500 (Server Error)!!1500.That’s an error.There was an error. Please try again later.That’s all we know.
• Error 500 (Server Error)!!1500.That’s an error.There was an error. Please try again later.That’s all we know.
• Error 500 (Server Error)!!1500.That’s an error.There was an error. Please try again later.That’s all we know.
• Knowledge of distributed architectures, asynchronous processing and event-driven architecture.
• Practical experience with AWS.
• Knowledge of AWS Lambda and serverless architectures.
• Knowledge of Docker and containerization concepts.
• Knowledge of CI/CD pipelines and deployment processes.
• Ability to put and maintain applications in production, not just develop locally.
• Ability to understand business requirements and transform them into functional products.
• Ability to investigate problems in a structured way and work with rapidly evolving technologies.

Specific AI knowledge

We expect practical knowledge in different aspects of modern AI application development, including:

• Large Language Models.
• Prompt Engineering.
• Context Engineering.
• AI Agents.
• Agentic Workflows.
• Function Calling/Tool Calling.
• Structured Outputs.
• RAG.
• Embeddings.
• VectorSearch.
• Model Context Protocol.
• Evals.
• LLM Observability.
• Guardrails.
• Human-in-the-loop.
• Context and memory management.
• Routing strategies between models.
• Token, latency and cost optimization.
• We do not necessarily expect prior experience with all existing frameworks. Knowledge of concepts and the ability to build and operate solutions are more important than mastery of a specific library.

Differences

• Experience with LangGraph, LangChain, Semantic Kernel, AutoGen, CrewAI or equivalent frameworks.
• Experience building MCP servers or clients.
• Experience with LLM observability and evaluation platforms.
• Experience creating internal tools to increase productivity of engineering teams.
• Experience with AI applications applied to the software development cycle.
• Knowledge of Spec-Driven Development and AI-assisted development workflows.
• Experience with Infrastructure as Code using Terraform, AWS CDK, CloudFormation or equivalent technologies.
• Experience with ECS, Fargate, EKS or other container launch models on AWS.
• Experience with event-driven architecture using SQS, SNS and EventBridge.
• Experience with enterprise environments and integration with corporate systems.
• Experience with Digital Commerce and platforms such as VTEX.
• Advanced or fluent English.

Expected profile

• We are looking for a person with a builder profile, capable of moving between product, engineering, infrastructure and Artificial Intelligence.
• This person must be able to receive a relatively open problem, understand the existing process, identify where AI can generate value, design the necessary architecture and transform this opportunity into a working product in production.
• We don't expect you to be a dedicated DevOps or Cloud Engineering specialist, but this person needs to have sufficient technical autonomy to understand how your application will be hosted, configured, monitored, scaled and operated, especially within the AWS ecosystem.

We also expect a critical stance regarding the use of AI. Not every problem needs an agent and not every workflow needs to be autonomous. The AI Product Engineer must know how to differentiate situations where traditional software, deterministic automation, LLMs or agentic architectures are the best alternatives.

More than knowing specific tools, we are looking for someone capable of deeply understanding the mechanisms behind these solutions, experimenting quickly and transforming successful experiments into reliable, observable, safe, measurable and sustainable products in production.

What you will find here

• An environment conducive to learning and professional growth 🎯
• Performance assessment and feedback, aiming for the continuous development of our people 📊
• Food and/or meal voucher for your grocery shopping and meals 🍴
• Medical and dental assistance to keep you and your family in good health 💙
• Agreement with pharmacies for discounts on medicines 💊
• Childcare assistance in accordance with current policy 🍼
• Partnership with SESC for varied cultural and leisure programs ✈
• Partnerships for your language studies, technology and course platform 📚
• Payroll loan with attractive rates + financial education program💰
• Corporate University and knowledge trails with diverse technology content, soft skills, market trends and much more 👨💻
• Referral Program with the possibility of prizes and bonuses 🎁
• Group life insurance ⛑