Senior AI Engineer

Description

Are you driven by technology, data, and innovation? At Triggo.ai, we transform complex challenges into intelligent solutions. We are a Brazilian company focused on Analytics, AI, and digital transformation, helping organizations make more strategic decisions and generate real business impact. We work with different market sectors, always combining cutting-edge technology, business vision, and agile execution. If you want to build solutions that make a difference and grow with a high-performance team, your place might be here. Come to Triggo.ai and build the future with us. **Responsibilities and attributes** * Design and implement multi-agent systems with tool orchestration, memory, routing, and fallback mechanisms; * Build and optimize RAG pipelines: ingestion, chunking, indexing strategies, hybrid search, reranking, and retrieval quality evaluation; * Develop APIs and services in Python (FastAPI) to expose AI capabilities, with testing, versioning, and CI/CD; * Define and implement the evaluation layer: reference datasets, quality metrics, prompt regression, and automated tests before each release; * Instrument LLM observability: execution tracing, tokens, cost per request, error rate per tool, and latency per stage; * Optimize cost and performance: model selection by task, reasoning effort adjustment, caching, batching, and context reduction; * Integrate agents with legacy systems and corporate channels, handling authentication, rate limits, idempotency, and reprocessing. **Requirements and qualifications** * Proven experience in advanced-level Python development: typing, async, project structuring, testing, and engineering best practices; * Real experience deploying LLM applications to production; * Mastery of agent orchestration frameworks (LangGraph, LangChain, CrewAI, Semantic Kernel, or equivalent proprietary SDK); * Practical experience with end-to-end RAG and vector databases or search engines (Azure AI Search, OpenSearch, pgvector, Qdrant, Elasticsearch); * Solid understanding of applied prompt engineering, structured outputs, function calling, and hallucination control strategies; * Experience with at least one public cloud in production, including containers, serverless, messaging, and identity and access management; * Familiarity with Docker, Git, and CI/CD pipelines; * Ability to communicate technically with clients, including presenting architecture decisions and defending trade-offs. **Differentiators** ---------------- We highly value those who have already built AI solutions on the AWS stack: * **Amazon Bedrock:** model invocation, Knowledge Bases, Guardrails, Agents, and model selection by use case; * **Amazon SageMaker:** inference endpoints, open model hosting, and training or fine-tuning jobs; * **Compute and orchestration:** EKS, ECS/Fargate, Lambda, and Step Functions for asynchronous and long-running workflows; * **Search and data:** OpenSearch Service (including vector and hybrid search), S3, DynamoDB, Aurora PostgreSQL with pgvector; * **Integration and events:** API Gateway, SQS, SNS, EventBridge; * **Governance and operations:** IAM with least privilege, VPC and private endpoints, Secrets Manager, KMS, CloudWatch, and X\-Ray for tracing. We are the **SysMap Group**, formed by the brands **SysMap Solutions, TriggoLabs, and triggo.ai**, an ecosystem of high-impact companies in the market, with innovative technology solutions that transform businesses and develop people. Our essence is driven by **passion for innovation, collaboration, and continuous growth**. For more than **25 years**, we have exceeded expectations by solving complex business challenges, operating in different segments and actively following technological evolution. Here, we believe that **technology only generates value when it is built by people**. Therefore, we cultivate an ethical, collaborative, and diverse environment where learning is constant and excellence in what we do is part of our way of being. **The only way to participate in SysMap Group selection processes is through the company pages on the Gupy platform, and there is no participation fee or hiring charge.**