Conversational AI
summary
Practical roadmap for conversational AI at Charles Schwab level: decide build versus buy based on strategic value, expected time to market and required NLP bench; start by mining call logs, FAQs and using clustering and verb‑noun pairs to produce intents, prioritize by traffic and integration complexity, and treat scaling thresholds as meaningful (<50, 50–200, >200 intents). Train with cleaned utterances, synthetic paraphrases, separation and cohesion scoring and k‑fold validation, deploy with a knowledge graph, monitor precision/recall and iterate — crawl before you walk.
AI-written summary · about
chapters (19)
- 0:00Introduction
- 2:34Importance of Chatbot
- 3:13How it works
- 5:44Consumers prefer humans
- 6:06Words
- 7:54Three Decision Points
- 10:40How do we get started
- 11:03Who owns the AI
- 11:52Other key players
- 13:27How do we go
- 13:42Poll
- 14:13Intent Instrumentation
- 15:45Prioritize Intents
- 17:14Model Training
- 19:41K4 Cross Validation
- 20:28K4 Deployment
- 21:17Performance Evaluation
- 22:33Product Analytics
- 23:26Learnings
related talks
01h 04mBuilding Expertise Into Generated Conversations
01h 04mBuilding Expertise Into Generated Conversations
30mBuild Conversational AI Interfaces with a Low Code Platform
20mNatural Language Interfaces for SaaS: Transforming Workflows with ConversationalAI
31mConversational AI Product Management
47mBuilding sophisticated conversational AI experiences
55mChatBots Were Yesterday, Conversational Commerce Is Today
39mHow Chatty Leads Builds AI Agents Businesses Actually Want