Senior AI Engineer (LLMs & Agentic AI)

This position is listed on behalf of a partner company, which manages all applications and next steps. Our partner is looking for a Senior AI Engineer (LLMs & Agentic AI) based in Brazil. Join a collaborative, AI-driven engineering environment focused on building and scaling production-grade AI applications. In this role, you will design and develop advanced large language model (LLM) solutions, agentic AI systems, and intelligent conversational workflows that deliver meaningful business impact. You will work with modern technologies from leading LLM providers, building reliable integrations, orchestration capabilities, and evaluation frameworks. The position offers significant technical ownership, with the autonomy to make architectural decisions and collaborate directly with engineering leadership. You will help improve AI performance, reliability, and efficiency through experimentation and engineering best practices. This is an opportunity to shape production AI capabilities within a fast-moving team that values technical excellence, clear communication, and continuous learning.

Build, maintain, and enhance Model Context Protocol (MCP) servers, tool definitions, and context management capabilities, including personally developing MCP servers from the ground up. Design and implement intent classification, evaluation, and routing workflows for natural language applications, ensuring requests are directed to the appropriate APIs and data sources. Develop multi-step orchestration workflows that coordinate multiple tools, APIs, and data sources, aggregating results into reliable responses. Integrate hosted large language models from providers such as Anthropic and OpenAI into production environments. Design and manage prompt engineering practices, prompt versioning, fallback strategies, and optimization initiatives to improve performance, latency, and cost efficiency. Develop and maintain LLM evaluation frameworks to measure model quality, validate behavior, and monitor production performance. Write and optimize SQL queries against cloud data warehouses to analyze model inputs and outputs and support evaluation and monitoring. Support backend functionality for conversational interfaces and MCP-enabled applications. Collaborate with software engineers and AI specialists to deliver robust, production-ready AI solutions. Participate in architectural discussions and contribute to the design and evolution of AI platforms. Investigate existing codebases, including partially documented legacy systems, and improve them through structured analysis and engineering best practices. Continuously experiment with new techniques, evaluate outcomes, and optimize AI systems for reliability, scalability, and maintainability. Communicate and defend technical decisions clearly, demonstrating ownership of implementation details and the ability to explain individual contributions in depth. Requirements: Senior-level experience building and deploying production AI applications using Python and large language model technologies. Strong proficiency in Python, software engineering principles, and best practices for building maintainable production systems. Strong SQL skills and practical experience working with cloud data warehouses. Hands-on experience developing production-grade LLM and agentic AI applications. Demonstrated experience personally designing and building an MCP server, rather than only integrating with or consuming an existing server. Experience implementing tool calling, context management, multi-step orchestration, and agentic workflows. Practical experience integrating hosted LLM APIs, particularly Anthropic and OpenAI, into production applications. Strong knowledge of prompt engineering, prompt versioning, fallback mechanisms, and strategies for optimizing model cost, latency, and performance. Experience designing conversational AI solutions, intent classification and routing systems, or multi-API orchestration workflows. Experience developing LLM evaluation methodologies and frameworks to assess model quality and production behavior. Ability to analyze, understand, and work effectively within partially documented or brownfield codebases. A bachelor's degree in Computer Science, Software Engineering, Artificial Intelligence, Data Science, or a related field is preferred; equivalent professional experience is also acceptable. Strong analytical, problem-solving, collaboration, and communication skills. Professional English proficiency sufficient to participate in technical interviews, explain complex implementation details, and sustain in-depth technical discussions with precision. Nice to have: Experience deploying, fine-tuning, and evaluating open-source models using Hugging Face. Nice to have: Familiarity with GPU acceleration, model inference optimization, and performance tuning. Nice to have: Experience with MLOps practices, including CI/CD, observability, and reproducible pipelines. Nice to have: Experience with vector databases, Retrieval-Augmented Generation (RAG), and applied natural language processing (NLP). Nice to have: Experience building client-facing AI products and scalable AI-powered solutions. Benefits: The position is based in Brazil; the specific work arrangement and remote-work eligibility are not specified in the original job description. Opportunity to work on production-grade LLM applications, agentic AI systems, and advanced conversational AI solutions. Significant technical autonomy and ownership of architectural and production design decisions. Collaboration with engineering professionals and AI specialists in a technology-focused environment. Opportunities to contribute to AI platform design and the development of scalable, AI-powered products. A learning-oriented environment that encourages experimentation, continuous improvement, and modern engineering practices. Salary, healthcare coverage, paid leave, bonuses, and other specific employment benefits are not specified in the original job description.