About the role

Senior AI Analyst working on advanced ML systems in partnership with the Brazilian Air Force. Atech develops critical technology solutions for defense and security.

Responsibilities

Architecture, deployment and optimization of AI models, with a focus on Machine Learning (ML) and Large Language Models (LLMs);
Development and fine-tuning of AI models using advanced RAG (Retrieval-Augmented Generation) techniques, LoRA, quantization and distillation;
Integration of AI models into production pipelines, ensuring scalability, security and computational efficiency;
Monitoring and maintenance of models in production, applying MLOps to manage the model lifecycle;
Collaboration with backend development teams, ensuring interoperability between AI models and Java-based systems and service-oriented architectures;
Data analysis and engineering, ensuring data quality and curation for AI model training.

Completed Master's degree in Artificial Intelligence, Computer Science, Data Engineering or related fields;
Advanced English, with the ability to communicate technically and write scientific content;
5+ years of experience in Machine Learning and LLMs, including architecture, tuning, deployment and optimization of models;
Proficiency with AI frameworks and libraries such as TensorFlow, PyTorch, Hugging Face Transformers and LangChain;
Experience with MLOps using Kubeflow, MLflow, Airflow or Vertex AI for automating and monitoring production models;
Data pipelines for AI, including feature engineering, data augmentation and handling imbalanced datasets;
Generative models and NLP, with in-depth knowledge of transformer-based architectures (GPT, BERT, T5, Llama, Mistral);
Experience with model optimization techniques, including quantization, sparsity, LoRA and supervised fine-tuning;
Integration of models into APIs and services using FastAPI, Flask or gRPC to serve models in production;
Experience with scalable inference architectures, including deployment on GPUs, multi-node environments and edge computing;
Versioning and reproducibility of experiments using DVC, Weights & Biases or MLflow;
Experience with vector databases such as FAISS, Pinecone, Weaviate or Milvus for implementing RAG (Retrieval-Augmented Generation);
Strong knowledge of cloud infrastructure for AI, with hands-on experience in AWS (SageMaker, Bedrock), GCP (Vertex AI) or Azure ML.

Meal and food allowance - Swile credit card;
Transportation allowance;
Gift voucher and birthday day off;
Bradesco medical and dental plan - to care for you and your family (direct dependents);
Gympass (Wellhub) – to support your physical health and energy;
Health and Wellbeing Program: +Saúde;
Quick massage at the office;
Childcare assistance for children up to 71 months;
Hybrid work model and flexible working hours;
Profit Sharing (PLR - Participation in Profits and Results);
Payroll-deductible loan options;
Private pension plan;
Life insurance.