Delivering the promise of conversational AI with Connect
summary
TrueServe shows a contact-center system that combines deterministic flows with probabilistic NLUs, demoing a 401(k) assistant that answers fact queries with 93–97% confidence and pulls facts from a Neo4j knowledge graph (one document yields 529 nodes). Pipeline uses Cognizer + embeddings, One Reach, Amazon Connect/Lex and a new DNA design/analytics layer, measures containment, abandonment and agent transfers, and supports human-in-the-loop approvals for taskful actions like payments or catch‑up contributions.
AI-written summary · about
chapters (17)
- 0:00Introduction
- 2:36Meet Daniel
- 4:04Agenda
- 5:04Building conversational AI today
- 8:59The dirty little secret of conversational AI
- 11:14Probabilities
- 15:33Demo
- 21:39Discovery Process
- 26:14Youre Never Done
- 29:10Trueserve
- 30:59DNA
- 33:53Interaction Patterns
- 37:45Knowledge Graphs
- 42:22More Questions
- 45:17Different Approaches
- 47:10Natural Language Processing NLP
- 50:00Human Loop Process
related talks
26mSuper-powered customer experience using Amazon Connect
38mAI-Powered Conversational Interfaces
31mSeamless Customer Service with Amazon Connect & Alexa
32mIncrease productivity and satisfaction with an intelligent contact center
47mBuilding sophisticated conversational AI experiences
46mReimagining clinical engagement using Amazon Connect
42mFrom AI Conversations to Data-Driven Decisions: Unlocking the Future of Customer Engagement
54melevating customer experience w/ conversational AI