As we reflect on Amazon IXD – VGT2, our annual gathering for industry leaders, we celebrate the remarkable advancements in technology that were unveiled. Over three weeks, we listened to insightful presentations and I am excited to share key announcements that are particularly impactful for the retail industry. These new offerings are designed to enhance store operations, create customer-centric experiences, improve demand forecasting, and elevate analytics capabilities.
Streamlining Edge Infrastructure
I recently authored a blog post about retail edge architecture that delves into strategies for executing compute and machine learning (ML) models at edge sites. This year’s event highlighted significant updates in this area, including AWS Panorama and 1U AWS Outposts servers.
AWS Panorama is an ML appliance coupled with a software development kit (SDK) that empowers retailers to implement computer vision (CV) on existing cameras located in their stores, distribution centers, and other venues. This technology allows for local, low-latency predictions with impressive accuracy. Previously, deploying CV required extensive processes, but AWS Panorama simplifies this by facilitating model deployment and camera stream discovery. For instance, Parkland Fuel, a Canadian fuel retailer, utilizes this device to derive insights regarding traffic and wait times at their gas stations. Retailers can either source CV models from AWS and our partner ISVs or develop their custom models using Amazon SageMaker, a fully managed service that allows rapid building, training, and deployment of ML models both in the cloud and at the edge.
The 1U and 2U AWS Outposts servers are compact, rack-mountable appliances tailored for environments like retail stores, where space is limited. These servers enable retailers to consolidate their applications onto a single managed device, streamlining IT operations and providing a foundation for deploying new applications from a centralized infrastructure.
Providing Location-aware Experiences
The Amazon Location Service equips application developers with geolocation capabilities, facilitating innovative experiences for shoppers. Enhance your buy-online-pickup-at-store (BOPUS) strategy with geo-fenced stores that automatically alert staff to prepare and deliver orders as customers arrive. Incorporating location awareness into customer engagement strategies can lead to new omni-channel opportunities, such as notifying customers about items in their online carts as they approach stores that stock those items. For a demonstration of Amazon Location Service, be sure to check out this blog post.
Enhancing Store-level Demand Planning
Amazon Forecast leverages the same technology that powers Amazon.com, providing developers with a fully managed service that employs ML for accurate forecasting, eliminating the need for extensive ML expertise. Now, with the Amazon Forecast Weather Index, forecasting has become even more accessible, as it integrates weather data to enhance predictions concerning foot traffic, product sales, and staffing needs.
Reducing Analytics Costs
Retailers often grapple with the challenge of consolidating data from multiple sources to gain a comprehensive view of their operations and customers. This typically necessitates an ETL (extract, transform, load) process, which can be costly and inefficient without proper optimization. AWS Glue Elastic Views, a new fully managed service, allows users to create materialized views from various cloud data stores, such as Amazon DynamoDB, Amazon S3, and others. This service continuously tracks changes in source data and automatically updates the materialized views in real-time, all while being serverless and scaling based on demand.
In the realm of customer support, Amazon Connect, the AWS contact center as a service, has rolled out several features designed to decrease call times and enhance agent productivity. With Amazon Connect Voice ID, customers can opt to authenticate their identity using their voice, thereby expediting the process. Amazon Connect Tasks simplifies the management of agent tasks with pre-built connectors for systems like Zendesk and Salesforce, creating a streamlined experience. Additionally, Amazon Connect Customer Profiles aggregates customer data from disparate sources, reducing research time, and Amazon Connect Wisdom leverages ML to offer search capabilities and real-time recommendations for agents. For further insights, this resource is excellent.
With additional announcements anticipated from Amazon IXD – VGT2, I encourage you to explore the latest updates through various sessions. Look out for more insights as the event progresses from January 12 – 14, 2021.
Leave a Reply