Creating Ethical Generative AI for Enterprises
summary
Hands-on demo of IBM Watsonx.ai shows building enterprise chatbots that use Watson Assistant plus Watson Discovery for OCR/KB search and RAG with embeddings (miniLM) to avoid hallucinations; models used include IBM Granite, Llama 2 and open-source Flan, with prompt tuning in prompt lab and dataset tuning in Tuning Studio. A helpdesk example classifies and auto-responds thousands of emails for a 10,000-user customer, summaries shorten handoffs to agents, and IBM offers a 2–4 week pilot to produce an MVP.
AI-written summary · about
chapters (12)
- 0:00Ethical generative AI intro
- 2:09Enterprises need AI chatbots
- 7:15Trustworthy AI challenges
- 8:44IBM Watson conversational AI
- 15:40Watson Discovery for unstructured data
- 16:01Introducing Watson X AI
- 23:18Enterprise AI use cases
- 27:44Trustworthy AI architecture
- 30:13Live demo Watson Assistant
- 39:04Prompt engineering demo
- 1:00:20Text classification & RAG demo
- 1:18:18Summary & next steps
related talks
01h 03mGenerative AI for Businesses
12mGoverning AI Use Cases with watsonx
48mScale and accelerate the impact of generative AI with watsonx
20mCreating Trusted AI for Business
14mArtificial Intelligence in Services
40mIBM Watson Conversation
25mHow AI Is Changing the Way We Work & Learn
54mIntroduction to Chatbots