Marketing Personalization · Real-time Analytics
Instacart needed real-time visibility into in-store shopping behavior, shelf conditions, and basket changes so retailers could reduce shrink, improve recommendations, and personalize physical-store experiences that traditional store systems could not observe well.
Reported impact
Instacart scaled Caper Cart deployments across more than 100 cities with 3x year-over-year growth, measured more than 1 percent incremental sales lift from in-store recommendations, cut whole-page ranking latency by 65 percent, and increased click-through rates by more than 5 percent.
Instacart turned its Caper Cart into a physical-AI retail edge platform by splitting the workload between on-cart sensor fusion and cloud ranking services. On each cart, NVIDIA Jetson Orin NX modules fuse three basket-facing camera streams with weight sensors and location inputs to build a real-time view of what the shopper has picked up and where the cart is in the store. That state feeds recommendation and reminder experiences, while cloud-side models running on NVIDIA Hopper GPUs use NVIDIA Dynamo to serve ranking and ad workloads faster than prior CPU deployments. The same architecture reuses signals from Instacart's online commerce graph to personalize in-store experiences, improve shelf visibility, and scale a single vision-language model across stores with minimal retraining.
Cart-mounted cameras, weight sensors, and location signals are fused on Jetson Orin NX modules to understand basket contents and shopper context in real time. Those signals are combined with Instacart's commerce graph in cloud inference services on Hopper GPUs and Dynamo, which return ranked recommendations, reminders, and retail analytics back to the cart experience.
By deploying NVIDIA Jetson for real-time sensor fusion, Instacart has turned a standard shopping cart into an AI platform that personalizes every trip and gives retailers visibility they've never had before.
By incorporating ranking signals from Instacart's online platform-drawn from 1.6 billion lifetime grocery orders-into in-store recommendations, Instacart measured more than 1% incremental sales lift in rigorous A/B testing.
Migrating ad and recommendation workloads from CPU to GPU cut whole-page ranking latency by 65% and item ranking latency by 40%, while click-through rates increased by more than 5%.
Scaled Caper Cart deployments across 100+ cities with 3x year-over-year growth, using a single vision-language model that serves all stores with minimal store-specific retraining.
Review how this reference was scored and trace each normalized claim back to its documented source.
Scoring methodology
Customer Reference Value measures usefulness as a customer reference. Evidence Confidence measures the quality and traceability of available evidence, not independent verification.
Traceability
Field coverage shows which facts each source supports. Dates are shown only when documented by the source.
Related
Alibaba Cloud Sample Oct 13, 2026 TOPSPORTS (滔搏) TOPSPORTS wanted to scale enterprise AI agents beyond isolated pilots, but every new agent risked recreating the same knowledge, data, tools, and skills in separate silos. Business-analysis, membership, and marketing… business analysis time reduced to 20 minutes AgentCore · AgentLoop · MCP tools 
Google Cloud Sample Modified · Oct 9, 2026 Central Retail Food Replaced ~2,000 monthly manual Excel reports with a single source of truth on BigQuery through Looker dashboards Central Retail Food unifies its ecosystem on Google Cloud, moving from fragmented data to conversational… Addresses 50% to 60% of customer service queries with automated chatbots built on Conversational Agents BigQuery · Gemini · Gemini platform Google Cloud Sample Modified · Oct 9, 2026 ICA Previously, however, ICA’s fragmented technical environment was holding it back. Data was siloed across legacy platforms and SQL server environments, creating high maintenance costs and preventing teams from sharing… ICA Gruppen is a leading Swedish retailer, with a portfolio covering grocery, pharmacy and banking. Serving… BigQuery · Bigtable · Gemini 
