Explore text-generation FMs for top use cases with Amazon Bedrock
summary
Shows concrete Bedrock workflows and results: customer support chatbot resolves an order issue in eight messages; Meta Llama 2 13B converts many transcripts into a product/issue/resolution table; AI21 Jurassic‑2 Ultra pulls action items, owners and due dates from meeting transcripts; Anthropic Claude (including Claude Instant) outlines blog posts and flags contract non‑adherence from transaction history; Cohere Command extracts parties and due dates from contracts. Notes 20 models on Bedrock, context sizes 4,000–200,000 tokens, fine‑tuning improves results with ~1,000 examples, and built‑in guardrails, private link and no default data reuse.
AI-written summary · about
chapters (20)
- 0:00Introduction and overview
- 1:08Top text generation use cases
- 2:55Customer support examples
- 5:46Meeting transcript analysis
- 7:13Blog post generation
- 8:17Legal document processing
- 10:48Available text generation models
- 12:15Choosing the right model
- 14:48Model training and tuning
- 16:34Language and context support
- 19:28Fine-tuning and customization
- 20:51Model pricing and selection
- 22:38Demo: Movie ad creation
- 29:51Prompt engineering best practices
- 34:57Monitoring and iteration
- 35:17Customer example: Booking.com
- 36:24Data privacy and security
- 37:53Guardrails for responsible AI
- 41:08Getting started resources
- 41:48Q&A and closing
related talks
43mBeyond Chatbots: How Amazon Bedrock Agents Drive Enterprise AI
38mCraft your own Generative AI Chatbot with Amazon Bedrock
31mLamRAG: From Data to Constructive Insights using Amazon Bedrock
29mAnthropic's Claude in Amazon Bedrock
28mBuilding database queries with Amazon Bedrock
24mBuilding GenAI applications with Amazon Bedrock
20mPower Amazon Bedrock applications with Neo4j knowledge graph
29mCustomizing genAI applications for your business