Kaleidoscope - Elevating Customer Experience
Customers are actively seeking ways to innovate through data and AI but are struggling to visualise end to end journey. This cross-industry demo uses a daily life use case to demonstrate how it could be implemented using AWS modern data tools, GenAI and AWS IoT. After seeing this demo, they would be able to apply similar architecture in their own innovation journeys.
In this episode of "Let's Talk About Data" our guest Shoaib Bustani is discussing Kaleidoscope, an innovative IoT solution that transforms customer feedback collection into actionable business insights using AWS services. The device addresses the challenge of capturing real-time customer sentiment at events and converting it into meaningful analytics that organizations can act upon immediately.
Kaleidoscope comes in three hardware versions built on Raspberry Pi: a light version with simple feedback buttons (green/amber/red), an advanced version with screens, and a touchscreen version for detailed questionnaires. The architecture leverages AWS IoT Core for secure device connectivity, with data flowing through S3, Lambda for enrichment, and EventBridge for aggregation. AWS Glue jobs analyze the data while Amazon Bedrock (Claude 3.7) generates AI-powered insights. The processed data is visualized through Amazon QuickSight dashboards, combining traditional BI with generative AI insights. The solution includes voice feedback capability using Amazon Transcribe and Comprehend for PII redaction, and features offline storage that automatically syncs when connectivity returns.
The solution has been successfully deployed at major AWS events including the London Summit, re:Invent, and Paris Summit, demonstrating its scalability across various industries from retail to financial services. Key insights from deployments are helping organizers schedule important content strategically. The project will be released as open-source under AWS Samples, allowing the community to adapt it for their specific use cases.
Key Highlights:
- Three Device Versions: Light (buttons only), advanced (buttons + screens), and touchscreen (detailed questionnaires)
- Near Real-Time Processing: Feedback processed within seconds through automated AWS pipeline
- Robust Architecture: IoT Core → S3 → Lambda → Glue → Bedrock → QuickSight for end-to-end analytics
- AI-Powered Insights: Claude 3.7 generates actionable recommendations from aggregated feedback data
- Offline Capability: Local storage with automatic sync ensures no data loss during connectivity issues
- Voice Feedback: Amazon Transcribe and Comprehend enable audio feedback with automatic PII redaction
- Cross-Industry Scalable: Adaptable to retail, manufacturing, financial services, and event management
- Open Source: Will be available in AWS Samples for community implementation
Hosts of the show 🎤
Ibrahim Emara - Database Specialist SA @AWS
Guests 🎤
Shoeb Bustani - Senior Enterprise SA @AWS
Enjoyed reading this content? Let the author know!
Your likes, comments, shares, and saves help creators reach more builders.
Loading recommendations
Loading article