Computer Vision Engineer (AI & Retail Analytics)

About Aravita At Aravita, we are on a mission to solve one of the world's biggest challenges: fresh food waste. Our AI-powered platform uses intelligent forecasting and order recommendations to optimize fresh food operations, reduce overstocking, and minimize out-of-stock items. We are a passionate, innovative team looking for a new member to help us build this fresh-focused technology and be part of a positive change. The Role We are looking for a highly-motivated Computer Vision Engineer to join our AI team. This is a hands-on role where you will be responsible for building, training, and deploying the computer vision models at the heart of our next-generation retail analytics platform. Your primary mission will be to analyze images from grocery store shelves to perform: SKU & Planogram Recognition: understand what products are where on the shelf. Quantity Estimation & Stockout Detection: accurately count or estimate product quantities and automatically detect empty spaces. This role is perfect for a problem-solver and a quick learner who is obsessed with shipping code and seeing their models create real-world impact. The outputs you build will be a critical input for our core forecasting and logistics models. What You'll Be Doing Build & Train: design and train deep learning models for object detection, image classification, and segmentation in a complex, real-world retail environment. Critically Think & Evaluate: you won't just build—you'll analyze. You will benchmark and integrate custom models (e.g., YOLO) against third-party AI services (like Google Vertex AI and Azure AI Vision) to find the best-fit solution for the problem. Master the Data: work with large, imbalanced, and often challenging image datasets and develop clever strategies for data augmentation and annotation. Collaborate & Deploy: work closely with our data science and engineering teams to deploy your models into a scalable production environment. Our Ideal Candidate Is a problem-solver first . You are excited by a complex challenge and have the critical thinking skills to break it down. Is a fast learner . You might not know every tool on day one, but you have the fundamentals and drive to learn them quickly. Is hands-on and results-oriented . You have a bias for action and a passion for building and shipping solutions. Has strong Python skills and hands-on experience with a major ML framework like PyTorch or TensorFlow . Has a solid understanding of computer vision principles and deep learning models (e.g., CNNs, object detection, transfer learning). What You'll Work With (And Learn) This is a fantastic opportunity to get hands-on experience with a state-of-the-art stack. You don't need to be an expert in all of these today, but you should be excited to learn them. Custom Models: YOLO, ResNet, and other SOTA architectures. Core Toolkit: Python, OpenCV, PyTorch/TensorFlow. Cloud AI Services: Google Vertex AI Vision & Azure AI Vision. Emerging Tech: exploring multimodal LLMs for image analysis. MLOps: production model deployment, monitoring, and versioning. Bonus Points (Nice to Have) A personal or academic portfolio of computer vision projects. Experience with managed AI services from Google Cloud or Azure. Familiarity with MLOps tools and best practices. Location: candidates based in or near São Paulo, SP , are preferred for occasional team meetups.

Location: São Paulo, BR

Posted Date: 11/15/2025