Leveraging Digital Twins and OpenSearch for Personalized Customer Experiences
summary
Builds a customer digital twin by generating PySim simulated shoppers and streaming synthetic and real events to S3 where Lambda pseudonymizes data and indexes OpenSearch vectorized events, then runs parallel sandboxes that call a RAG pipeline to generate actions like targeted discounts. Uses weekly/monthly grouping of time series to create personas, swaps in time‑aware LLMs for prediction, and claims sandbox experiments converge with real customers over time, lowering cost and compliance risk while enabling faster A/B testing.
AI-written summary · about
chapters (8)
- 0:00Personalization and digital twins
- 4:45Digital twin use cases
- 11:13Synthetic data for personalization
- 17:58Challenges in ML models
- 20:27OpenSearch for vector search
- 26:30Time-series chunking
- 33:19Key takeaways and compliance
- 36:09Q&A
related talks
38mUser Behavior Insights
32mPersonalisation and Specialisation of Search With OpenSearch Agentic Search
21mPersonalize Search Results with OpenSearch Agentic Memory
43mOpenSearch & AI Driven Cloud Native Platforms: Adaptive & Agentic Microserv...
26mHyper-Personalized Marketing with Digital Platforms & Generative AI
29mMachine Learning in Production: What to do after you’ve trained the models?
46mDelivering the next wave of innovation with OpenSearch
26mRevolutionizing Search: Unveiling Conversational Search