RETAIL AI

Intelligence That Transforms Retail

Deploy real-time analytics across every customer touchpoint while reducing infrastructure complexity. When advanced AI meets retail innovation, extraordinary customer experiences become standard.

Try Metis Today

Read more

Unlock the Future of Shopping

Transform every retail touchpoint with AI that works where it matters most. The Axelera® AI Metis® delivers advanced computer vision capabilities for fraud prevention at checkout, real-time inventory monitoring, and personalized digital signage that adapts to individual shoppers. With 214 TOPS processing power and class-leading energy efficiency, retailers achieve faster response times while maintaining customer privacy through local processing. The Voyager SDK simplifies AI vision system deployment with YAML-based pipeline configuration that works with pre-trained models.

Try Metis Today

Or talk to us first

60%

of sales are digitally influenced.

This number is expected to grow as AI personalizes recommendations, helps streamline decision-making and manages auto replenishment.¹

52%

of respondents have increased capital allocation for loss prevention.

Retailers see the value of AI and computer vision applied to loss prevention technologies from self-checkout monitoring to behavioral anomalies and more.²

23%

of customers are lost to competitors when retailers have outdated inventory systems.

Computer vision with AI can monitor shelf status, track product movement, and predict demand patterns in real-time.³

AI Powered Customer Experience

Smarter stores. Happier Shoppers.

Your customers expect convenience, speed, and personalization. Computer vision at the edge with AI delivers that without delays, outages, or cloud dependency. With Axelera AI's Metis, you get unprecedented price-performance ratio that allows system manufacturers to create massive opportunities at costs that don't limit innovation.

Comprehensive store coverage without complexity is possible because Metis gives you the ability to stream 24 concurrent 1080p video feeds.

Consistent accuracy is critical for real world retail environments, which is why in-memory computing delivers where reliable operation matters most.

Try Metis Today

Or talk to us first

The Future of Retail Uses Computer Vision AI

AI-enabled Point of Sales

Computer vision monitors self-checkout to reduce loss prevention in scenarios such as label switching. Object detection networks ensure that scanned items match their labels, reducing errors and enhancing user experience. Using video recognition instead of weight and scale measurements enables self-checkouts to become more efficient and user-friendly.

However, the high costs associated with current hardware solutions, which often rely on expensive GPUs, hinder large-scale implementation. This financial barrier often results in innovations getting stuck in the pilot phase, preventing widespread adoption.

Purpose built hardware combines the power of the cloud with the power budget of edge devices all while working with standard camera infrastructures. Reducing energy costs while spurring innovation lowers the total cost of ownership.

Car with Shopping Bags

Optimize shopping with new insights

Smart video systems monitor shopper traffic and queues in supermarkets and malls, optimizing staff shifts to reduce wait times and improve the shopping experience.

Additionally, computer vision can recognize car license plates, streamlining curbside pickup processes and improving reliability. This ensures orders are accurately matched with arriving vehicles, reducing errors and enhancing customer satisfaction.

These advancements showcase how computer vision AI-driven solutions can transform retail efficiency and customer service.

Enable large scale autonomous stores

In retail environments, combining object detection, classification, and pose estimation is crucial for tasks like automated inventory management, customer behavior analysis, and enhancing shopping experiences. For example, a system can use object detection to identify products, classify items, and pose estimation to understand customer interactions with products on shelves. The Axelera AI Metis design, with its four self-sufficient cores, efficiently handles these tasks in parallel, significantly enhancing performance.

By utilizing existing camera infrastructures, Axelera AI's solution avoids the substantial investment required to replace cameras with expensive AI-equipped alternatives. This makes it feasible to scale such solutions across large stores. Moreover, Axelera AI’s hardware brings data center-level performance into a compact and cost-effective form factor, reducing the total cost of ownership and enabling widespread deployment in retail settings.

The exceptional performance and accuracy of the Axelera AI Acceleration platform have significantly fueled our collaborative efforts. Its unmatched performance-to-price ratio, surpassing traditional GPUs and dedicated AI Processing units, has been critical in our selection process. We are confident that leveraging their state-of-the-art YOLO performances will empower us to tackle new challenges in our current and future video analysis applications.

Alexandre Perez, R&D Director at XXII

Q1. How do AI accelerators differ from GPUs in edge retail applications?

AI accelerators such as the NPU or AIPU are specialized hardware designed to speed up tasks related to artificial intelligence whereas GPUs process a broad range of computing tasks and are general purpose. For edge computing such as computer vision, AI accelerators offer greater efficiency in terms of increased performance and power savings.

Q2. How much power does an AI accelerator consume compared to a GPU?

Axelera AI's Metis AIPU uses just 11.5W to reach a performance height of 214 TOPS while a comparable GPU can run up to 60W to get TOPS in the 200s. Performance per watt is significantly more efficient with an NPU.

Q3. How fast can a retailer implement NPU-based systems?

The speed of implementation varies greatly based on the scope and complexity of the system, but Axelera AI's Metis saves time and money compared with alternatives because it allows you to use existing camera infrastructure. The Voyager SDK includes a comprehensive model zoo, making ease of deployment greater.