Event: Artificial intelligence meets search: Powering a new generation of insights
April 24, 2024 |
seedling
Event #
ref AI Powered search Recorded event. Sponsored by . “AI will fundamentally change every software category, starting with the largest category of all – search,” said Satya Nadella, Chairman and CEO, Microsoft.
tags :
SCHEDULED: #
Attendees[0/0]
#
Preparation #
Agenda #
Notes #
Knowledge management and Knowledge discovery is the biggest category for LLM application.
Vector Store in RAG is a way to reduce hallucination by injecting the right context. Everyone is becoming knowledge worker in one form or another.
Questions during the meeting #
Actions #
TODO #
- apply this on bank statements
OCR of Images #
2024-01-10_10-16-33_screenshot.png #
2024-01-10_10-24-52_screenshot.png #
Haley Sutherland Research Manager, Conversational AI & Intelligent Knowledge Discovery IDC Mandy Andress CISO Elastic Alicia Frame Principal Product Manager, Azure Open AI Microsoft
2024-01-10_10-20-53_screenshot.png #
The Al Everywhere Chapter in the Digital Business Era Business Value Digital Business: Innovation at Scale A new chapter in the Digital Business Era is starting now - Al Everywhere. 2035 Generative Al triggered the opening of this new chapter because it seeks to drastically reducing the time and long-term costs associated with developing solutions across a wide range of use cases associated with automation and intelligence. Use Case Use Case Use case Widening UseCase AI Use Case Use Case UseCase 9 Use Case Use Case Generative Use Case AI Use Case AI Everywhere Use Case Use Case PET - AR LLMS APIs Use Case This chapter will be about how we use data as an input and as a business outcome. Use Case Narrow AI 15 Cloud Use Case Xaas Use Case Platforms & Communities Internet Use Case Social Use Case Use Case Multiplied Innovation Mobility 2010 Time CIDC OIDCI 2
2024-01-10_10-22-42_screenshot.png #
Generative Al will be a key component of knowledge management and knowledge discovery applications going forward based on the popularity of GPT-3.5, GPT-4 and other GenAl LLMS What Generative AI use cases do you anticipate having the most promise for your organization? Knowledge Management. Applications 46% Conversational. Applications 46% Design Applications 44% Code Generation Applications 43% Marketing Applications 36% 0%6 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% Source: Future Enterprise Resiliency & Spending Survey Wave 6, IDC, July 2023, N=890 CIDC o IDC 4
2024-01-10_10-27-12_screenshot.png #
Despite improvements over the years, the process ofsearching across and combining data from multiple sources still represents the biggest waste oftime for knowledge workers. There are many things that can waste time related to search & knowledge discovery. As a user of internal/enterprise search, how many hours a week do you spend on each of the following activities? Average Hours Lost to Search Activities Manually combining data from multiple sources 3.0 Searching across multiple: systems for the: same information 2.5 Answering FAQS because users cannot find information 2.3 Recreating content that can't be found 1.3 Searching for, buti not finding, information .1 Re-sending content that cant be found 1.0 Translating from one language to another 0.9 1.00 0.00 0.50 1.50 2.00 2.50 3.00 3.50 n=! 522 Source: North America Knowledge Discovery Survey, IDC, February, 2023 EIDC o IDCI 6
2024-01-10_10-28-02_screenshot.png #
Retrieval Augmented Generation (RAG) Vector-Based Search Provides Pre-Response Augmentation of LLM with Specific Data enterprise data (process query as embeddings) Query (vector) database D * context USER Response* (vector) search/retrieval best fit snippets/docs synthesis & generation LLM *provide additional grounding for user e.g. footnotes, source links, etc prompts CIDC IDC 7
2024-01-10_10-39-13_screenshot.png #
Leading organizations are investing in the latest Al-powered knowledge discovery technologies to modernize infrastructure, gain new insights across data, people, process, and culture, and improve efficiency and productivity. You indicated that you have been using your current internal and/or external search software for 3 years or less. What were the initial motivations to implement or upgrade a search/knowledge discovery system? Modernize <nowledge infra astructure 41.6 Gain new insights across data, people, process, and culture 39.4 Streamline work processes and improve efficiency/productivity 38.1 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% (% of respondents) n =231 IDC Source: North America Knowledge Discovery Survey, IDC, February, 2023
OCR of Images #
2024-01-10_10-16-33_screenshot.png #
2024-01-10_10-24-52_screenshot.png #
Haley Sutherland Research Manager, Conversational AI & Intelligent Knowledge Discovery IDC Mandy Andress CISO Elastic Alicia Frame Principal Product Manager, Azure Open AI Microsoft
2024-01-10_10-20-53_screenshot.png #
The Al Everywhere Chapter in the Digital Business Era Business Value Digital Business: Innovation at Scale A new chapter in the Digital Business Era is starting now - Al Everywhere. 2035 Generative Al triggered the opening of this new chapter because it seeks to drastically reducing the time and long-term costs associated with developing solutions across a wide range of use cases associated with automation and intelligence. Use Case Use Case Use case Widening UseCase AI Use Case Use Case UseCase 9 Use Case Use Case Generative Use Case AI Use Case AI Everywhere Use Case Use Case PET - AR LLMS APIs Use Case This chapter will be about how we use data as an input and as a business outcome. Use Case Narrow AI 15 Cloud Use Case Xaas Use Case Platforms & Communities Internet Use Case Social Use Case Use Case Multiplied Innovation Mobility 2010 Time CIDC OIDCI 2
2024-01-10_10-22-42_screenshot.png #
Generative Al will be a key component of knowledge management and knowledge discovery applications going forward based on the popularity of GPT-3.5, GPT-4 and other GenAl LLMS What Generative AI use cases do you anticipate having the most promise for your organization? Knowledge Management. Applications 46% Conversational. Applications 46% Design Applications 44% Code Generation Applications 43% Marketing Applications 36% 0%6 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% Source: Future Enterprise Resiliency & Spending Survey Wave 6, IDC, July 2023, N=890 CIDC o IDC 4
2024-01-10_10-27-12_screenshot.png #
Despite improvements over the years, the process ofsearching across and combining data from multiple sources still represents the biggest waste oftime for knowledge workers. There are many things that can waste time related to search & knowledge discovery. As a user of internal/enterprise search, how many hours a week do you spend on each of the following activities? Average Hours Lost to Search Activities Manually combining data from multiple sources 3.0 Searching across multiple: systems for the: same information 2.5 Answering FAQS because users cannot find information 2.3 Recreating content that can't be found 1.3 Searching for, buti not finding, information .1 Re-sending content that cant be found 1.0 Translating from one language to another 0.9 1.00 0.00 0.50 1.50 2.00 2.50 3.00 3.50 n=! 522 Source: North America Knowledge Discovery Survey, IDC, February, 2023 EIDC o IDCI 6
2024-01-10_10-28-02_screenshot.png #
Retrieval Augmented Generation (RAG) Vector-Based Search Provides Pre-Response Augmentation of LLM with Specific Data enterprise data (process query as embeddings) Query (vector) database D * context USER Response* (vector) search/retrieval best fit snippets/docs synthesis & generation LLM *provide additional grounding for user e.g. footnotes, source links, etc prompts CIDC IDC 7
2024-01-10_10-39-13_screenshot.png #
Leading organizations are investing in the latest Al-powered knowledge discovery technologies to modernize infrastructure, gain new insights across data, people, process, and culture, and improve efficiency and productivity. You indicated that you have been using your current internal and/or external search software for 3 years or less. What were the initial motivations to implement or upgrade a search/knowledge discovery system? Modernize <nowledge infra astructure 41.6 Gain new insights across data, people, process, and culture 39.4 Streamline work processes and improve efficiency/productivity 38.1 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% (% of respondents) n =231 IDC Source: North America Knowledge Discovery Survey, IDC, February, 2023