Showing posts with label AI. Show all posts
Showing posts with label AI. Show all posts

Saturday, December 21, 2024

Thoughts about Artificial Intelligence and business value

The question that a lot of companies have, is  how to realize business value from AI. Obviously companies don't invest millions (billions) of dollars in artificial intelligence just to play around with LLMs. Companies need to quantify the business value they want to realize.

To get an answer to this question, you might want to take a look at the white paper  Research paper - Artificial Intelligence and business value: a literature review (2021, Ida Merete Enholm, Emmanouil P., Patrick Mikalef and John Krogstie) which provides a framework for connecting AI with business value.

The white paper distinguishes between first and second order effects of GenAI. First order effects are the immediate, direct impacts of implementing Generative AI whereas second order effects are indirect, longer-term impacts that result from the first-order effects. 

While first order effects (e.g. enhancing business decisions, increasing process efficiency and transforming business processes) are crucial for justifying the initial investments and demonstrating immediate ROI, second order effects (impact on operational and financial improvements, sustainability etc...) are essential for long term strategic planning and understanding the broader implications of Generative AI.



Saturday, September 28, 2024

Dynamics 365 and Power Platform monthly reading list September 2024

 2024 Release Wave 2

Release plans for 2024 release wave 2 are now publicly available - Release plans for Dynamics 365, Power Platform, and Cloud for Industry - Dynamics 365 | Microsoft Learn - from July 29th, early access will be available for activation - other key dates for 2024 release wave 2 are also available . Also check out the videos like Dynamics 365 Service 2024 Release Wave 2 or Dynamics 365 and Power Platform 2024 Release Wave 2 - these might also contain features/functionality which are not listed in the release plans.

Copilots, AI and machine learning

Technical topics (Configuration, customization and extensibility)
Topics for Dynamics 365 Business Applications Platform consultants, project managers and power users

Sunday, April 21, 2024

The ABC of AI: Retrieval-Augmented-Generation (RAG) and grounding

This is the first in a series of blog posts about more advanced generative AI and Large Language Model (LLM) concepts which I use as notes to myself (check out Why I blog and you might want to consider it as well)

Retrieval-augmented generation (RAG) is an AI framework that enhances the quality of responses generated by large language models (LLMs). LLMs are trained on a massive amount of data and understand statistical relationships between words but lack true comprehension of their meanings. So when faced with specific questions in a dynamic context so that is where RAG comes in.

RAG integrates information retrieval into LLM answers by using these steps:

  1. User inputs prompt: when you ask a question, RAG uses your input prompt
  2. RAG retrieves relevant information from an external knowledge base based on the user prompt
  3. RAG combines this external content with your original promt creating a richer input for the LLM
RAG and grounding are related concepts in the context of enhancing LLMs. Grounding is the process of providing LLMs with information about a specific use-case. RAG is one of the techniques which is used for grounding (another technique is dense retrieval - see Dense X Retrieval: What retrieval granularity should we use  for more details). 

Microsoft Copilot offers a good example of grounding and RAG in use. The reason why Copilot is able to give more targetted responses, is because it uses grounding to improve the specificity of the prompt. 

Copilot uses Microsoft Graph, which can retrieve information about relationships between users, activities and organizational data (like info in Power Platform/Dataverse and/or Dynamics 365, info from e-mails, chats, documents and meetings) as part of the prompt grounding process. Microsoft Copilot will use the user prompt and additional info retrieved through Microsoft and then sends it to the LLM. For more details see Microsoft Copilot for Microsoft 365 overview


The most common systems to provide external data for RAG LLMs are vector databases and feature stores.

References:

Monday, April 15, 2024

Dynamics 365 and Power Platform monthly reading list April 2024

Copilots, AI and machine learning

 

Technical topics (Configuration, customization and extensibility)


Topics for Dynamics 365 Business Applications Platform consultants, project managers and power users


Wednesday, November 29, 2023

Dynamics 365 and Power Platform monthly reading list November 2023

 2023 Release Wave 2

Technical topics (Configuration, customization and extensibility)

Copilots, AI and machine learning

Topics for Dynamics 365 Business Applications Platform consultants, project managers and power users


Friday, July 21, 2023

Dynamics 365 and Power Platform monthly reading list July 2023

2023 Release Wave 2

Technical topics (Configuration, customization and extensibility)

Topics for Dynamics 365 Business Applications Platform consultants, project managers and power users


Friday, January 14, 2022

Dynamics 365 and Power Platform monthly reading list January 2022

2021 Retrospective

Technical topics (Configuration, customization and extensibility)

Topics for Dynamics 365 Business Applications Platform consultants, project managers and power users

Thursday, April 02, 2020

Dynamics 365 and Power Platform monthly reading list February 2020

Dynamics 365 and Power Platform - 2020 Wave 1 Topics
Technical topics  (Configuration, customization and extensibility)

Monday, January 14, 2019

Whitepaper– Artificial Intelligence in Belgium and Luxembourg: Outlook for 2019 and beyond


Microsoft recently commissioned a study with over 277 European companies on how companies currently manage their AI activities and how they see the challenges and opportunities ahead – 21 companies out of Belgium and Luxembourg have participated in this study amongst them Ageas, Afga Gevaert, Alpro, Arcelor Mittal, Katoe Natie, KBC Group, Proximus and Solvay – some interesting findings:
  • Machine learning (81%), smart robotics (62%) and NLP (48%) are most useful for companies in Belgium and Luxembourg. Smart robotics ranks considerably higher than the European average (44%)
  • Current use cases highlighted by executives include predictive maintenance, using demand forecasting to optimize supply chain operations and automation of logistics
  • Advanced analytics is considered to be one of the most important capabilities (out of 8 capabilities) necessary for AI success in Belgium and Luxembourg
Download AI in Belgium and Luxembourg: How 277 major European companies benefit from AI

Thursday, December 06, 2018

LearnAI: Enterprise Knowledge Mining bootcamp training material

I recently stumbled upon this interesting training material https://azure.github.io/LearnAI-KnowledgeMiningBootcamp/ . It explains how to built an enterprise search solution by applying knowledge mining to business documents like contracts, memos, presentations and images and expose these using a bot.  You will use Microsoft Azure Cognitive Search, Azure Bot Framework, Azure Functions, etc..

Tuesday, November 06, 2018

Speaking at Dynamics 365 Saturday Brussels 2018

I will be delivering a session about chat bots and Dynamics 365 at Dynamics 365 Saturday Brussels on November 17th. Registration is free. Dynamics 365 Saturday is a Technical & Strategy event organized by the community with the following credo: “Learn & share new skills while promoting best practices, helping organizations overcome the challenges of implementing a successful digital transformation strategy with Microsoft Dynamics 365”.

Wednesday, October 24, 2018

Dynamics 365 monthly reading list October 2018

Technical topics (Configuration, customization and extensibility)

Topics for Dynamics 365 CE business analysts, project manager and users