Senior AI Engineer

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

📋 Description

• Design and create multi-agent systems featuring tool orchestration, memory management, routing, and fallback mechanisms.

• Construct and enhance RAG pipelines, which encompass ingestion, chunking, indexing, hybrid search, reranking, and retrieval quality assessment.

• Develop APIs and services utilizing Python (FastAPI) to make AI capabilities accessible, incorporating testing, versioning, and CI/CD practices.

• Define and execute the evaluation layer using benchmark datasets, quality metrics, prompt regression, and automated testing prior to each release.

• Implement LLM observability metrics, including tracing, token consumption, cost per request, error rates per tool, and latency at each step.

• Enhance cost and performance via model selection, tuning reasoning efforts, caching, batching, and context reduction techniques.

• Integrate agents with existing legacy systems and corporate channels, addressing authentication, rate limits, idempotency, and reprocessing requirements.

⛳️ Requirements

• Demonstrated experience in advanced Python development, including typing, asynchronous programming, project structuring, testing, and engineering best practices.

• Practical experience in deploying LLM applications in a production environment.

• Expertise in agent orchestration frameworks such as LangGraph, LangChain, CrewAI, Semantic Kernel, or similar proprietary SDKs.

• Hands-on experience with end-to-end RAG processes and vector databases/search engines like Azure AI Search, OpenSearch, pgvector, Qdrant, and Elasticsearch.

• Solid understanding of applied prompt engineering, structured outputs, function calling, and strategies for mitigating hallucinations.

• Experience working with at least one public cloud in a production capacity, including containers, serverless architectures, messaging, and identity and access management.

• Familiarity with Docker, Git, and CI/CD pipelines.

• Ability to engage in technical communication with clients, including presenting architectural decisions and justifying trade-offs.

• Plus: experience with Amazon Bedrock, Amazon SageMaker, EKS, ECS/Fargate, Lambda, Step Functions, OpenSearch Service, S3, DynamoDB, Aurora PostgreSQL with pgvector, API Gateway, SQS, SNS, EventBridge, IAM, VPC, Secrets Manager, KMS, CloudWatch, and X-Ray.

🏝️ Benefits

• Remote work opportunities.

• Position is also open to candidates with disabilities.