Brazil
Power Digital
ArcelorMittal
Mid-Level AI Engineer
Brazil
Belo Horizonte, Minas Gerais, Brazil
PythonSQLDockerKubernetes
Automatic translation. Check the original.
About the opportunity
Hello! We're glad you're interested in one of our opportunities.
How does our selection process work?
Our selection process is conducted in a structured, transparent, and respectful manner, considering both technical aspects and cultural fit.
Throughout the journey, we aim to understand not only your experiences but also how you connect with environments that value:
- Safety
- Collective care
- Responsibility in decisions
This is the initial stage of the process. As you advance, other stages include interviews with the People team and leadership, as well as complementary assessments throughout the process and, at the end, the hiring stage.
We are very happy with your interest in joining our #TIMAÇO and hope to receive your application. To learn more about us, visit our social media.
See you soon!
#SejaArcelorMittal
At ArcelorMittal, we believe that a career choice goes beyond the position; it's about being part of an environment that values people, promotes continuous development, and, above all, places Safety as a non-negotiable value in all decisions.
Below, you'll find more details about ArcelorMittal, the opportunity, what we offer, and what we look for in people who want to build this journey with us.
Who We Are?
We are ArcelorMittal, the largest steel producer in Brazil and a global company that operates with the purpose of producing intelligent steels for people and the planet.
We are an organization that continuously evolves, driven by people, innovation, and the constant pursuit of more efficient and sustainable solutions. ArcelorMittal is one of the world's leading integrated steel and mining companies, with operations in 60 countries and primary steel production in 15. With over 125.000 employees worldwide, we serve customers in various sectors, such as automotive, construction, energy, and engineering. Our products are part of everyday life, making a positive impact on individuals, communities, and society.
In our culture, people are at the center of everything we do, which is why Safety, our fundamental value, guides our actions, decisions, and the way we work every day.
Being part of ArcelorMittal means joining an environment in constant transformation, where:
- Safety and care for people always come first
- Innovation and technology are applied to generate value responsibly
- Learning is continuous, with real challenges and growth opportunities
- Diversity of perspectives strengthens our decisions and results
We seek people who want to grow with the company, contributing responsibly, collaboratively, and committed to sustainable results.
You Will Be Responsible For
- Designing and maintaining serving pipelines in AKS, ensuring versioning, rollback, and high availability.
- Designing and implementing production RAG with incremental ingestion, corpus versioning, and automatic update policies.
- Developing and evolving AI agents for production, integrating LLMs with APIs, tools, and corporate systems.
- Defining and versioning technical contracts for tools.
- Optimizing inference flows and creating semantic caching mechanisms to reduce cost and latency.
- Implementing end-to-end observability (tracing, latency, cost, errors, and throughput).
- Managing cost metrics per route and tenant, implementing alerts and budgets in the pipeline.
- Conducting automatic and human quality evaluations (factuality, safety, and usefulness), including guardrails.
- Automating tests and validations before promotion to production and standardizing the deployment of models, prompts, and policies.
- Ensuring complete traceability of models and agents via MLflow 3.x.
- Implementing access control and auditing (tokens, secrets, and logs).
- Collaborating with the semantic team and the agent team to ensure grounding and integrity of calls.
- Documenting architectures and technical decisions and participating in code reviews.
- Supporting the development of junior profiles, disseminating best practices.
- Ensuring that activities in your area align with Health and Safety requirements;
- Knowing and Complying with the golden rules and FPS requirements.
Projects You Will Work On
- Strategic HR and Controller Projects: participation in initiatives relevant to the business, focusing on transforming data into insights and supporting strategic decisions.
- Building the Data & Analytics area: opportunity to contribute to the definition of standards, processes, and best practices, helping to structure and strengthen the area.
- Generating business impact: working closely with partner areas, connecting business needs with analytical solutions that generate value.
- Development opportunity: experience that combines business vision, analytical capacity, and participation in building solutions and processes.
Minimum requirements
Minimum Requirements
- Bachelor's degree in Computer Science, Engineering, Data Science, AI, or related fields.
- Hands-on experience in developing and integrating generative AI solutions in production (LLMs, RAG, and agents).
- Advanced Python for AI applications (asynchronous APIs, automation, testing, and monitoring).
- LLMs, prompt engineering, and context engineering.
- Production RAG: ingestion, chunking, embeddings, vector databases, semantic search, and continuous updating.
- Agent and orchestration frameworks (LangChain, LangGraph, LlamaIndex, or Databricks Agent Framework).
- Containers and cloud orchestration: Docker and Kubernetes/AKS (Helm, autoscaling, canary deploy, rollback).
- CI/CD, Git, and deployment of AI applications in production.
- MLflow 3.x (tracking, registry, evaluation, and auditing) and Unity Catalog.
- Observability of LLMs and agents (OpenTelemetry, tracing, latency, cost, errors, and throughput).
- Cost and performance management of LLMs (semantic cache, cost metrics per route/tenant, alerts, and budgets).
- Evaluation of LLMs and agent quality (automatic metrics and human evaluation of factuality, safety, and usefulness).
- REST APIs, technical contract design, and integration with corporate systems and SQL/NoSQL databases.
- Security and governance for AI solutions (access control, tokens, secrets, audit logs, and guardrails).
- Automated testing and software engineering best practices (code review, technical documentation).
It will be a plus if you have:
- Azure OpenAI / Azure AI Foundry, Amazon Bedrock, or Vertex AI.
- MCP (Model Context Protocol), multi-agent systems, and GraphRAG.
- Tools for evaluating LLMs and agents (Langfuse, LangSmith, Ragas, DeepEval, or similar).
- Notions of red teaming and AI Gateway.
- Machine Learning and Deep Learning.
- Agile methodologies (Scrum).
Benefits
At ArcelorMittal, we take care of people holistically, promoting health, well-being, safety, and quality of life for you and your family.
To take care of your physical and mental health:
- Platform to contribute to your health and well-being care;
- Medical and Dental Health Plan (extendable to legal dependents)
- Mental health care line and live psychology
- Our commitment to Health and Safety is reflected in continuous initiatives that prioritize people's physical, mental, and emotional well-being.
Special For You and Your Family
- Childcare Assistance
- Food Allowance
- Vacation Return Bonus
For Your Future
- Private Pension Plan
More opportunities with one profile
Want to be considered for more roles without repeating the whole process? Create your profile and be found by companies using Nort.
Create my profile