PstrongOverview /strong /ppbr/ppYour role will be based in Vitoria-Gasteiz (Hybrid) to make key contributions to AI projects with a primary focus on computer vision, alongside natural language processing, foundational models and other general AI/ML solutions aligned with PEP AI roadmap within the global digital transformation journey within PepsiCo. /ppbr/ppYou will be part of a collaborative interdisciplinary team around data and analytics, where you will explore new technologies, and design the adoption of new products in the domain of computer vision AI to support the development of digital solutions to create business decision making. Your work will span retail analytics, quality control systems, supply chain optimization, and consumer insights through visual intelligence. On the one hand, you will work closely with process owners, product owners and final business users. This will provide you the correct visibility and understanding of criticality of your developments to business stakeholders and ST functional leaders. On the other hand, you will work closely with other data scientists to transfer your findings to ensure the scalability, robustness and reusability of the solutions that you develop. /ppbr/ppYou will be an internal ambassador of the team's culture around data and analytics. You will provide stewardship to colleagues in the solution development landscape and further lead the onboarding of new developments in AI within the PepsiCo domain leading the development of best practices and standard ways of working. /ppbr/ppstrongResponsibilities /strong /pulliActive contributor to code development in projects and services related to computer vision, object detection, image segmentation, video analytics, machine learning, mathematical modeling, AI and ML/AI Operations, and deep learning. /liliDesign and deploy CV models for real-world applications including retail execution monitoring, manufacturing quality control, supply chain optimization, and visual search. /liliActive contributor in innovation activities, exploring the boundaries of technology to create AI solutions with the ability to continually prove accuracy and with a design goal of productization/reusability. /liliCreate documentation for learnings and knowledge transfer. /liliCreate reusable packages or libraries. /liliPartner with data engineers to ensure data access for discovery and proper data is prepared for model consumption. /liliPartner with other data scientists working on industrialization. /liliOccasionally, coordinate work activities with business teams, other services and as required. /liliCommunicate with data scientists and other ST stakeholders in the process of service design, training and knowledge transfer. /liliSupport large-scale experimentation and build analytical models. /liliSet KPIs and metrics to evaluate analytics solution given a particular use case. /liliRefine requirements into modelling problems. /liliInfluence product teams through data-based recommendations. /liliResearch in state-of-the-art methodologies. /li /ulpbr/ppstrongQualifications /strong /pulliMaster's Degree in Computer Science, Data Science, Mathematics, Statistics, Computer Vision, or a related quantitative field. /liliPreferred: PhD in Computer Science, Machine Learning, Computer Vision, or related fields, with specialization in deep learning for visual recognition. /liliMinimum: 4+ years of hands-on experience building, deploying, and evaluating computer vision models in a research or industry setting, with a strong focus on object detection, segmentation, or image classification applications. /liliIdeal: 6+ years of experience, with demonstrable success in developing computer vision solutions, and delivering tangible business value. /liliComputer Vision models and architectures. /liliExperience with CNNs (ResNet, EfficientNet, MobileNet), Vision Transformers (ViT, Swin), object detection models (YOLO, Faster R-CNN, DETR), and segmentation models (U-Net, Mask R-CNN, SAM). /liliUse cases: object detection, image classification, semantic/instance segmentation, visual quality inspection, OCR, video understanding, etc. /liliDeep Learning Frameworks: Proficiency in PyTorch, TensorFlow, or similar frameworks. /lili8+ years working in Python, Java or other programming languages /liliPython: Essential for data manipulation, model development, and integration. Data science libraries: NumPy, Pandas, Scikit-learn, Matplotlib. Computer Vision libraries: OpenCV, Pillow, Albumentations. Deep learning: PyTorch Vision, TorchVision, timm. Detection frameworks: Ultralytics, MMDetection, Detectron2. /liliExperience with model optimization and deployment: ONNX, TensorRT, quantization, edge deployment. /liliData annotation tools: CVAT, Label Studio, or similar. /liliCloud Computing: Experience with platforms like AWS, Azure, and GCP. /liliAzure DevOps: Knowledge of CI/CD pipelines and software development practices using Azure DevOps. /liliFluency in English: Strong written and verbal communication skills in English are essential for collaboration and conveying technical ideas. /liliHigh proficiency in business storytelling and communicating insights in business consumable format. /liliExperience with multimodal AI (vision-language models like CLIP, BLIP, GPT-4V) is a plus. /liliExperience with Responsible AI and FAIR data is a plus. /liliExperience with NLP and generative AI for language tasks is a plus. /liliTrack record of Developing AI products that have generated business value, and impact across multiple Industries. /liliStrong problem-solving mindset and creatively to identify and address challenges and opportunities /liliStrong written, oral communication, and interpersonal and leadership skills, ideally experience driving consensus /liliStrong project management, organizational, and prioritization skills with the ability to deal with ambiguity while juggling multiple priorities /liliDeep passion for AI technologies and their real-life applications /liliAbility to think strategically within business context and work at a fast pace /li /ul