Normalized claim
Accuracy: 25% decrease
Inventory forecasting accuracy improved by 25%, helping procurement maintain proper stock levels and reduce shortages or excess inventory.
Supr Daily Improves Delivery Verification and Inventory Management with Amazon Rekognition and Amazon Forecast
Use case typeComputer vision checkoutUpdated Jun 13, 2026
Supr Daily, a grocery ordering and delivery service based in Bangalore, serves over 200,000 customers with fresh groceries delivered early morning across six cities. The company faced challenges with manual verification of delivery photos which was slow, error-prone, and led to unnecessary refunds, as well as manual inventory forecasting inefficiencies. To scale and improve, Supr Daily implemented AWS machine learning services: Amazon Rekognition Custom Labels for automated image verification with 95% accuracy and 350 ms latency, and Amazon Forecast for demand forecasting that improved inventory management accuracy by 25%. The solution automated delivery verification in near real-time, reducing manual work and enabling instant feedback to delivery partners to improve photo quality and reduce fraudulent refund claims. Inventory management became simpler and more accurate through automated forecasting and notifications sent to procurement, helping ensure timely stock replenishment. The backend infrastructure is hosted on AWS Elastic Beanstalk to support scalability for millions of customers. Overall, the implementation supports a 70% user growth and enhances partner and customer satisfaction.
- Organization
- Supr Daily
- Industry
- Retail
- Location
- India
- Published
- April 2026
Reported outcomes
+70%
quantified impactOther quantified impact
−25%accuracy
Strategic outcomes
New product / capabilityAutomated delivery photo verificationBetter decisions & insightImproved inventory demand forecastingSpeed & agilityEnabled near real-time delivery feedbackScale & capacitySupported scalable growth across cities
Why do we believe this?Outcome claims, sources, and evidence checks
Normalized claim
Quantified impact: 70% increase
The system scales seamlessly to support a 70% increase in user base and thousands of delivery partners across multiple cities.
Why do we believe this deployment?Customer identity, provider attribution, maturity, and source checks
- Customer
- Supr Daily
- Provider
- AWS
- Maturity
- Unknown
- Linked source
- AWS Customer Stories
No explicit deployment-stage evidence found.
Customer identity supportedSource describes one deploymentMaturity evidence evaluated
Primary read
Use case focus
Showing 3 of 3
- 1Image recognition
- 2Inventory forecasting
- 3Delivery verification
- Manual delivery photo verification was slow, error-prone, and caused unnecessary refund payouts.
- Manual inventory forecasting was inefficient and hindered scalable growth.
- The growing user base and delivery volume required a more scalable and automated system.
- Implemented Amazon Rekognition Custom Labels to automate near real-time delivery photo verification with 95% accuracy, reducing manual checking workload.
- Used Amazon Forecast to automate inventory demand forecasting, improving accuracy by 25%, and integrated notifications via Amazon Simple Queue Service (SQS) to procurement teams for timely stock management.
- Built the infrastructure on AWS Elastic Beanstalk for scalable hosting and ease of iteration without managing server infrastructure.
- Leveraged Amazon S3 for storing images and data for model training and operations.
- Delivery verification process accelerated to 350 milliseconds with high accuracy, enabling instant feedback and reducing refund errors by eliminating manual review bottlenecks.
- Inventory forecasting accuracy improved by 25%, helping procurement maintain proper stock levels and reduce shortages or excess inventory.
- The system scales seamlessly to support a 70% increase in user base and thousands of delivery partners across multiple cities.
- Partner and customer satisfaction improved due to faster, more reliable delivery confirmation and inventory availability.
Architecture
Supr Daily uses Amazon Rekognition Custom Labels for ML-based automated image verification of delivery photos with 95% accuracy and 350 ms latency. Images are stored in Amazon S3, and Amazon Forecast automates demand forecasting for inventory management. Notifications are sent to procurement teams using Amazon SQS. The application backend is hosted on AWS Elastic Beanstalk to ensure scalability.
Sources & evidence1
Evidence: Medium50/100Evidence strength
- Customer explicitly identified
- Primary source available
- Quantified outcome available
- Technical implementation details available
Type: Customer StoryPublished: Apr 29, 2026Publisher: AWS Customer StoriesEvidence: PrimaryConfidence: High
AI-generated summary. Verify important details with the linked sources before relying on this case.
Explore related AI use cases
Customer evidence
Provider evidence
AI patterns
Browse the catalog
Was this useful?
Community
Comments
No published comments yet.