DBA Basics: Getting Started with Performance Tuning.pdf
Reading the IBM AI Strategy for Business
1. 1
Reading the AI strategy for
Business and digitization.
1
Pietro Leo
Executive Architect for Data, AI and Automation for large ISVs
IBM AoT/Open Innovation Community Member,
Head of Center for Advanced Studies
2. 2
2
2
By 2030, the global GDP will
grow by 14%, equivalent to
$15.7 trillion, thanks to
artificial intelligence1
• Half of the growth will come from
labour productivity savings, and the
other half from increased consumer
demand because of improvements in
AI-based products 1
• 60-70% of business task time could
be compressed using generative AI
models.2
• 62% of executives believe that
generative AI will disrupt the way
their organization designs and
shapes experiences and processes in
the company3
Source: 1) PwC; 2) McKinsey; 3) IBM IBV;
4. 4
4
4
+AI → AI+
IBM Strategy for AI for Business
Digital Transformation
Process/Task
Automation
New Hybrid
Processes
+
5. 5
5
5
The modern-day AI ladder
+AI
AI+
Collect, organize, grow data
Add AI to your applications
Replace/update your workflows
Automate tasks in your workflows
AI orchestrates the work to be done
6. 6
6
6
+AI
AI+
Collect, organize, grow data
Add AI to your applications
Replace/update your workflows
Automate tasks in your workflows
AI orchestrates the work to be done
All Data and Products and for kind of AI processing
Classic Algorithms + AI Models –
Enterprise AI layer
Business Task Automation
Workflow Re-Engineering e
Ottimizzazione -
Intelligent Process Layer
New Use-Cases,
Products, Services
The modern-day AI ladder
7. 7
7
7
Classic IT
Business
Processes
Data Depbt / Data Collection /
Accuracy / Costs
Business Tasks / Business Use Cases / Business Workflows
Training
from scratch
Fine Tuning
Retrieal & Search
Augmented
General Models
LLMs Apps (e.g. ChatGPT,
Co-pilots, etc)
e.g. Search / Answer Tasks
e.g. Scoring / Assessing Tasks
e.g. Summarization / Translation Tasks
AI
Models
Traditional AI
Generative AI
Enterprise AI layer
AI Enterprise layer: it is intended to host thousands of specialized AI models of different types, not just a single
large generalist model. These models will be derived from open models and adapted for specific use cases or
settings: Traditional models and generative models will be used in a hybrid fashion.
8. 8
8
Data Depbt / Data
Collection / Accuracy /
Costs
Business Tasks / Business Use Cases / Business Workflows
Fine Tuning
Retrieal & Search
Augmented
General Models
e.g. Search /
Answer Tasks
e.g. Scoring / Assessing Tasks
e.g. Summarization / Translation Tasks
e.g. Claim management workflow
Training from
scratch
Intelligent Process Layer
Classic IT
Business
Processes
AI
Models
Traditional AI
Generative AI
Enterprise AI layer
Intelligent Business Processes layer: dynamically orchestrates and optimizes the work of business
workflows, considering heterogeneous activities performed by both humans and autonomous systems.
9. 9
9
The industry’s leading unified
hybrid cloud app platform
watsonx
Enterprise-ready generative AI
and data platform
Open Targeted
Trusted Empowering
Based on the best AI and multi-
cloud and hybrid technologies
Designed for targeted business
use cases
Built with governance,
transparency, and ethics
Bring your own data and models &
run anywhere
IBM: Open
Innovation,
AI Tech
trustworthy
and
problem
solver for
enterprises
Consulting
Trusted solution provider for
enterprises
Business Focus
Customer Experience and Care
Employee Experience, Productivity and Talent
IT development, coding, app modernization, and
operations
Core business operations, sustainability
AI Competencies
20k+ Skilled data and AI
practitioners, +1k on
Consultants skilled in
generative AI
Enterprise Projects
~2K AI use cases developed
for clients, 40k+ Active and ongoing AI and analytics
client engagements worldwide
Industrialized frameworks to support the full
development cycle: from early PoC, to MVP, to scalable
production
10. watsonx
10
Hybrid cloud
AI tools
Data services
AI and data
platform
SDKs and APIs
AI assistants
Scale and accelerate the impact of AI with trusted data, an open architecture, and
seamless integration
watsonx Orchestrate
watsonx Assistant
watsonx Code Assistant
watsonx Orders
Build on a consistent, scalable
foundation based on open-
source technology
Empower individuals to do
work without expert
knowledge across a variety of
business processes and
applications
Use programmatic interfaces to
embed watsonx platform
capabilities in assistants and
applications
Leverage generative AI and
machine learning — tuned with
your data — with
responsibility, transparency
and explainability
Access data fabric services to
define, organize, manage, and
deliver trusted data to train
and tune models
Red Hat OpenShift AI
(e.g., Ray, Pytorch)
watsonx
watsonx.ai
watsonx.governance
watsonx.data
Ecosystem
integrations
Data fabric
services
Foundation models
Open Source | Hugging
Face
Llama 2 | Meta AI
Geospatial | IBM +
NASA
Granite | IBM
• Digital Labor
• IT Automation
• Security
• Sustainability
• Application
Modernization
• Data and
Transaction
Processing
IBM Software
Infrastructure
• IBM Z
• Distributed
Infrastructure
• Infrastructure
Support
• Quantum
AI+
Software
and SaaS
Partners
and ISVs
Multi - Public Clouds
AWS ● Azure ●
Others
Enterprise
Infrastructure
Edge
Open Innovation