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

Sunday, February 16, 2025

How will AI impact your IT budget?

As AI begins to make its way into the Enterprise, how will the experimentation, adoption and operations of AI impact 2025 and beyond IT budgets? 


SHOW: 898

SHOW TRANSCRIPT: The Cloudcast #898 Transcript

SHOW VIDEO: https://youtube.com/@TheCloudcastNET 

CLOUD NEWS OF THE WEEK: http://bit.ly/cloudcast-cnotw

CHECK OUT OUR NEW PODCAST: "CLOUDCAST BASICS"

SHOW NOTES:


WHERE WILL IT BUDGETS MOST LIKELY BE IMPACTED BY AI IN 2025-2026?

  • Is your focus on human-augmented capabilities or human-replacement capabilities? 
  • If you’re buying AI as part of SaaS services, will the pricing be seat-based (human-augmented) or usage-based (human-replacement), or some combination of both? 
  • Are you directly paying for AI-capabilities, or is AI being positioned as “free”, but the other services are becoming more expensive?
  • Are you expanding your predictive AI projects (and data-science teams), or primarily focused on GenAI related projects? 
  • Are you buying AI resources from the cloud providers, or from AI-specific hosting providers, or buying GPUs for private-cloud usage? 
  • Are you planning to use standard AI tools, or do you plan to training/align your own models with your data?
  • Do you have a cost-control capability for your AI projects yet?
  • What are your success-criteria for POCs, Experiments, Trials? 
  • Are you upskilling existing staff, or hiring AI-specific staff? Or using external consulting? 


FEEDBACK?



Sunday, September 1, 2024

Slow Data, Fast Data and Adapting AI models for your business

With so much rapid change happening around AI models, is there a framework for when to build models, train models, fine-tune models, use RAG or something else? Let’s explore…

SHOW: 852

SHOW TRANSCRIPT: The Cloudcast #852 Transcript

SHOW VIDEO: https://youtube.com/@TheCloudcastNET 

CLOUD NEWS OF THE WEEK:  http://bit.ly/cloudcast-cnotw

CHECK OUT OUR NEW PODCAST:  "CLOUDCAST BASICS"

SHOW SPONSOR:

SHOW NOTES:


HOW SHOULD WE DECIDE WHICH AI MODEL TO USE FOR OUR APPLICATION?

  • Build, buy, use open source, augment with your own data - so many decisions
  • How often do you update models with new data? 

HOW DO WE BRING OUR COMPANY DATA / KNOWLEDGE TO OUR AI MODELS?

  • Large Static Models - Predictive AI
  • LLMs + Slow-Changing Data - Align the Model
  • LLMs + Fast-Changing Data - RAG
  • LLMs + Agent-Learned Data - Agentic AI

FEEDBACK?