Η παρουσίαση Master the art of Data Science πραγματοποιήθηκε από τον Παντελή Ξανθούλη, Analytics Software Sales | IBM, στην εκδήλωση της εταιρίας μας, InTech – Accelerate your AI journey.
2. Key take-aways
The AI Ladder
AI is not magic, it’s a journey!
Data fuels digital transformation
AI unlocks the value of Data
Hybrid cloud democratizes Data
3. Predict and shape future outcomes
Empower people to do higher value work
Reimagine new business models
AI is shaping the future of work
Automate decisions, processes, experiences
How AI
pioneers
see value
28%
72%
Cost
Savings
Revenue
Increase
IBM & Inttrust
4. AI-powered
advertising
engagement
4
Predict fraud across
their web & mobile
banking system
Predict power
demand by for
renewable energy
Predict and target first-
time buyers in the US
Surface hidden
insights to optimize fantasy
football outcomes
Visually categorize
damage & instantly
issues quote
However, AI is not magic
Cognitive car manual
explaining increased vehicle
complexity
Achieved a 40% call
deflection rate with
virtual agents
Mercedes-Benz
Identifies gaps in
terms in complex
RFPs
Optimize cardiac care
in high volume remote
regions
Our learnings from
experience in helping
thousands of enterprises
put
AI to work iKure
IBM & Inttrust
5. 5
Turning AI aspirations into outcomes
DATA TALENT TRUST
The lifeblood of AI, but
complexity slows progress
60%
Are challenged in
managing data quality
AI skills are rare
and in high demand
62%
Are challenged to acquire
talent [and build skills]
Skepticism of AI systems
& processes
62%
Need an approach to
AI production readiness
find operationalizing, sustaining
and scalingAI challenging
Stuck in
Experimentation 51%
Based on 2019 Forrester “Challenges That Hold Firms Back From Achieving AI Aspirations”
IBM & Inttrust
6. 81%
6
No amount of AI algorithmic sophistication
will overcome a lack of data [architecture]“ Data collection & preparation is the most
time consuming and difficult part of AI
8x
more likely to
have a robust
data architecture
do not understand
the data needed for
AI
Sources: 2018 MITSlone ”Reshaping business with AI”
There is no AI
without an IA
AI pioneers are
[Information architecture]
IBM & Inttrust
7. COLLECT - Make data simple and accessible
ORGANIZE - Create a business-ready analytics foundation
ANALYZE - Build and scale AI with trust and transparency
INFUSE - Operationalize AI throughout the business
AI
MODERNIZE
Make your data ready for an
AI and hybrid cloud world
The AI Ladder
A prescriptive approach to the journey to AI
One Platform,
Any CloudTalent &
Skills
8. Cloud Pak
for Data
Speed digital transformation by digesting and unifying volumes of
data for real-time AI insights delivered as cloud AI microservices
Modernize
Make your data ready for an AI and hybrid cloud world
Virtualize all data,
regardless of where it lives
Dynamically scale on-demand to
accommodate changing needs
Integrate and govern data across
hybrid cloud and data settings
ONE unified set of cloud-native
data & AI services, on any cloud
Automate the end-to-end data
and AI lifecycle management
An open, extensible information
architecture for AI
IBM & Inttrust
9. Infuse
Operationalize AI throughout the business
Instill trust and transparency across
business leaders and users
Innovate with new business models
optimized by industry vertical needs
Leverage AI to streamline
knowledge work & productivity
Speed time to value with pre-built
apps (e.g., customer service)
Automate analytical planning, forecasting,
budgeting, etc.
Employ AI-assisted business
intelligence and data visualization
Using pre-built and custom AI to empower advisors to better know
clients and shape improved outcomes across 350K inquiries per day
Watson
Applications
& Solutions
IBM & Inttrust
10. 10
The Ladder to AI
IBM’s AI Portfolio
Everything you need for Enterprise AI, on any cloud
Watson
Knowledge
Catalog
Watson
Studio
Watson Machine
Learning
Watson
OpenScale
Build Deploy Manage
Interact with Pre-built AI Services
Watson Application Services
Catalog
Unify on a Multicloud Data Platform
IBM Cloud Private for Data
AI Open Source Frameworks
Watson Solutions
Health | Financial Services | Retail | etc
IBM & Inttrust
12. 3 Primary Use Cases Customer Care
Through the Watson Assistant, IBM
can decrease call center operations
cost, while improving the customer
experience and developing new
revenue streams
Conversational Commerce
Provide guided buying experience for
prospective customers to purchase
goods and services through the
mobile or messaging channel of their
choice
Employee Productivity
Simplify access to common questions
and tasks through enterprise channels
14. Watson Visual Recognition is An image recognition
service that enables users to
quickly and accurately tag,
classify, and train visual
content using machine
learning.
BASIL
LEAF
HERB
PLANT STEM
GREEN
What is Watson Visual Recognition?
15. Watson Visual Recognition focuses on
Assessment
Watson Visual Recognition
assesses for better problem-
solving.
What is Watson Visual Recognition?
Identification
Watson Visual Recognition
identifies objects and people.
Categorization
Watson Visual Recognition
categorizes for easy organization.
Recommendation
Watson Visual Recognition
recommends for faster decision-
making.
hatchback
compact car
vehicle
claret red color
vintage
modern
Fender bender, 87%
confident
Historically, we’ve paid
$7,500 for similar types
of damage
16. Why are enterprises struggling to
capture the value of AI?
Tools &
Infrastructure
• Need an
environment that
enables a “fail fast”
approach
• Discrete tools
present barriers to
productivity
Governance
• If the data isn’t
secure, self-
service isn’t a
reality
• Challenge
understanding
data lineage and
getting to a system
of truth
Skills
• Data Science skills
are in low supply
and high demand
• Nurturing new data
professionals is
challenging
Data
• Data resides in
silos & difficult to
access
• Unstructured and
external data
wasn’t considered
IBM & Inttrust
17. The building blocks of AI
• Find, Catalog, mask data
• Built in compliance
• Advanced transformation
capabilities
• Organize data so that it
can be trusted
• Open platform for Data Science
• Descriptive, predictive to
prescriptive
• ML deployment
• Analyze insights on demand
• All Sources of Data
• The Common Application
layer
• Write once, deploy
anywhere
• Relevant data and make
it simple & accessible
Collect Organize Analyze
2. Solution Overview
IBM & Inttrust
19. Search and Explore with Data Privacy 2. Solution Overview
IBM & Inttrust
20. Analyze any data, no matter where it lives
Connect to and analyze your data without moving a single
through dozens of connectors and multiple deployment
Empower your entire organization with notebooks,
visual productivity, and automation tools
Leverage your entire organization with a variety of tools in a
integrated platform
One platform to rule them all from discovery to
production
Analyze data, build predictive models, and seamlessly integrate
Watson Machine Learning to deploy
IBM Watson Studio
Enterprise Data Science platform that helps your
team work together to build models to make better
data driven decisions for your business
IBM & Inttrust
21. • Integrated with Watson Studio and
Watson Machine learning
• Automatically ingest, clean, transform, and
model with hyperparameter optimization
• Training feedback visualizations provide
real-time results to see model
performance
• One-click deployment to Watson Machine
Learning
21
IBM AutoAI
22. Cloud Pak for Data: Modular
Cloud-native Data Micro Services
Collect Data Organize Data Analyze Data
Data Virtualization In-Memory
Warehouse
Relational Database
NoSQL (MongoDB)
Data Visualization
Machine Learning
Text Mining/NLP
Watson
SPSS Modeler
Transformation
Profiling
Masking
Govern
Prescriptive Analytics
(Optimization)
Quality
Real Time
Streaming
ETL - DataStage
23. Cloud Pak for Data
Cloud Pak for
Data
Collect,
organize,
and analyze data
With Machine Learning
capabilities and AI Model
Deployment
IBM containerized
software
Container
platform and
operational services
Watson
APIs
Real Time
Streams
Mongo Cognos Watson Studio Machine / Deep LearningSPSS
On-Premise
Credit goes to Clay Davis
RHEL / Kubernetes Based
Appliance
Data virtualization
Data warehouse
Governance catalog & discovery
services
Data Integration services
Data Visualization & Dashboards
Data Science: Model Design &
Deployment
Collect Organize Analyze
Insights Platform
IBM & Inttrust
24. IBM & Inttrust
Consumer Layer
(Interface Provisioning)
Applications Mobile Apps
Analytics
Tools
Portals Web
Services
Virtualization Layer
Caching &
Optimization
Connection Layer
(Adaptors)
Governance Catalog
(Metadata)
Consumers
Data Sources
Warehouses
Marts
Cloud Applications
Web
Services
Lakes
Files
NoSQL
Virtualization
Platform
Data Virtualization
25. Life event and financial
event prediction
Predict life and financial events
impacting client’s financial lives
to help advisors proactively
service a client’s needs
Machine learning
accelerators to
provide insights
Dynamic segmentation
Advanced dynamic client
segmentation helps identify
unique cohorts of clients by
behaviors, account profile
information, and
demographics
Client attrition
Ability to predict client attrition at
configurable points in the future
to protect revenue and
wallet share, while also building
profound client loyalty
IBM & Inttrust
Offer Affinity
Run campaigns effectively
by identifying client product
propensities or investment
theme affinities to drive new
or more desirable business
opportunities.
Intelligent Maintenance for Asset
Intensive Industries
(Telecommunications, Manufacturing , Oil and Gas, and Transportation)
Use Machine Learning to calculate optimal maintenance day
Notas do Editor
We all know data is the foundation for businesses to drive smarter decisions. Data is what fuels digital transformation. But, it is Artificial intelligence (AI) that unlocks the value of that data, which is why AI is poised to transform businesses with the potential to add almost 16 trillion dollars to the global economy by 2030.
However, adoption has been slower than anticipated. Business leaders not only need to understand the power of AI, but how they can fully unleash its potential and operate in a hybrid, multicloud world.
This presentation aims to demystify AI, present common AI challenges and failures, and finally, provide a unified, prescriptive approach (which we call “the AI Ladder”) to help organizations unlock the value of their data and accelerate their journey to AI.
Across an array of use cases, AI pioneers are employing a core set of new AI capabilities and shaping the future of work. They’re leveraging AI to:
Predict and shape future outcomes
Empower people to do higher value work
Automate decisions, processes, and experiences
Reimagine new business models that are trusted, transparent, and deployable anywhere
Some of our clients paving the way are:
Geisinger (Predict and shape future outcomes) – One sepsis patient dies every two minutes in the US, but over 80 percent of deaths are preventable with prompt diagnosis and care. Geisinger partnered with IBM and used the data science tools available in IBM Watson Studio to develop machine learning models capable of analyzing thousands of patient records and medical journals. Working with IBM, Geisinger has successfully built a predictive model for sepsis mortality based on real-life EHR data, that has helped researchers identify clinical biomarkers that are associated with the higher rates of mortality from sepsis.
Woodside Energy (Empower people to do higher value work) - Woodside, Australia’s largest independent energy company has been a global leader in oil and gas for over half a century. Their secret? Hire and develop heroes. This formula has helped Woodside build some of the largest structures on the planet, in some of the most remote parts of the ocean, and safely transport the energy they produce to people around the globe. To ensure the next generation could successfully carry the torch, Woodside knew they had to harness the instinctual know-how of their best employees. This goal — to create a cognitive business to augment and share their tribal knowledge — is what led Woodside into an industry-first partnership with IBM and Watson.
Experian (Automate decisions, processes, and experiences) -- Experian collects and aggregates information on over one billion people and businesses including 235 million individual US consumers and more than 25 million US businesses. Currently they use a rules-based system to determine if a file should be automatically loaded or checked. However, this system misses may correct files, resulting in additional human labor to check files and load them manually. Experian worked with IBM Watson Studio to reduce the number of files sent to be checked by approving them directly in the load process using machine learning techniques to classify these documents and descriptive analytics to get a better understanding of how the current rules are working, and successfully reduced 96% of false positives and 95% of false negatives.
Legalmation (Reimagine new business models) -- With intuitive IBM® Watson® offerings, LegalMation developed a first-of-its-kind AI platform to automate routine litigation tasks. Supported by the IBM Watson ecosystem, the company quickly launched its solution for drafting early phase response documents, helping legal teams save time, drive down costs and shift strategic focus.
But those AI pioneers are just some of the few. There are thousands of clients IBM is helping put AI to work, and while these clients are on various stages in their own journey, they all have one thing in common -- they are transforming their business through the use of AI.
However, to ensure a successful AI strategy, organizations need to understand how to adopt and implement the technology and realize there will be failures along the way. In order to turn AI aspirations into outcomes, organizations need to overcome three major AI challenges: data complexity, skills, and trust.
AI Challenge: Data Complexity
Data is the foundation and fuel for businesses to drive smarter decisions particularly as they embark upon digital transformations. The problem is that while 90 percent of business leaders list improving the use of data as a top priority, only 15 percent of them are actually getting what they need from their data. As a result, the majority of businesses have a plan to build a system of insights to become data-driven and have declared the journey to AI as a strategic priority. Data is the lifeblood of AI, and if organizations don’t solve for its complexities, their progress can be slowed by data siloes, incomplete data, and the appropriate approach to governance.
AI Challenge: Skills
Data is the lifeblood of AI, but you also need skills , such as knowing how to code, understanding and building deep learning and machine learning models, to bring AI to fruition. The challenge is that AI skills are rare, and therefore in high demand, so there’s a shortage of skilled workers available to hire. This makes it even more important that the technology being built and used is more easily accessible to everyone within the business, regardless of skill level.
AI Challenge: Trust
For organizations to truly embrace and scale AI across the entire businesses, they need to break open the ‘black box’ of AI and trust the systems. It is critical to ensure AI recommendations or decisions are fully traceable – enabling enterprises to audit the lineage of the models and the associated training data, along with the inputs and outputs for each AI recommendation. As more applications make use of AI, businesses need visibility into the recommendations made by their AI applications. In the case of certain industries like finance and healthcare, in which adherence to GDPR and other comprehensive regulations present significant barriers to widespread AI adoption, applications must explain their outcomes in order to be used in production situations.
For AI to thrive, and for businesses to reap its benefits, it is imperative that organizations are able to address these three challenges to ensure they have trust their AI systems, have the right skills across their organization, and access to their data, no matter where it resides.
As companies look to harness the potential of AI, they need to use data from diverse sources, support best-in-class tools and frameworks, and run models across a variety of environments. However, 81 percent of business leaders do not understand the data and infrastructure required for AI.
According to MIT Sloan, “No amount of AI algorithmic sophistication will overcome a lack of data [architecture]…bad data is simply paralyzing.”
Put simply: There is no AI without IA (information architecture).
IBM recognizes this challenge our clients are facing. As a result, we’ve built a prescriptive approach (known as the the AI ladder), to help clients overcome these challenges and accelerate their journey to AI, no matter where they are on their journey. It allows them to simplify and automate how organizations turn data into insights by unifying the collection, organization and analysis of data, regardless of where it lives. By climbing the ladder to AI, enterprises can build a governed, efficient, agile, and future-proof approach to AI. Furthermore, it is an organizing construct to the Data and AI products and services offered by IBM and our business partners, and it is the technology foundation to unify how those products and services work together.
What we have learned from AI pioneers is that every step of the ladder is critical. AI is not magic and requires a thoughtful and well-architected approach. For example, the vast majority of AI failures are due to data preparation and organization, not the AI models themselves. Success with AI models is dependent on achieving success first with how you collect and organize data.
The AI ladder has four steps (often referred to as “rungs”):
Collect: Make data simple and accessible.Collect data of every type regardless of where it lives, enabling flexibility in the face of ever-changing data sources.
Organize: Create a business-ready analytics foundation.Organize all the client's data into a trusted, business-ready foundation with built-in governance, protection, and compliance.
Analyze: Build and scale AI with trust and explainability.Analyze the client's data in smarter ways and benefit from AI models that empower the client's team to gain new insights and make better, smarter decisions.
Infuse: Operationalize AI throughout the business.Operationalize AI throughout the business - across multiple departments and within various processes - drawing on predictions, automation, and optimization.
Then, spanning the four steps of the AI ladder is the concept of Modernize, which is how clients can simplify and automate how they turn data into insights by unifying the collection, organization and analysis of data, regardless of where it lives, within a multicloud data platform.
What is meant by "modernize"? Modernize means building an information architecture for AI that provides choice and flexibility across a client's enterprise. As clients modernize for an AI and multicloud world, they will find that there is less "assembly required" in expanding the impact of AI across the organization.
IBM has the depth and breadth of capabilities to help clients:
Deploy an information architecture for AI
Prepare data for a multicloud world
Infuse AI everywhere, with confidence
Harness the flexibility and growth of open source technologies
Speed time-to-value with unprecedented simplicity & agility
Our lead-with products supporting this rung are:
IBM Cloud Pak for Data
IBM Cloud Pak for Data Systems
Most organizations are still in the early days of determining how and where to use AI to advance their business agenda. But, interest is growing, largely because AI has the potential to solve one of the biggest challenges we face: we’re drowning in data – 2.5 quintillion bytes – generated every day, but we’re starved for insights. AI helps us take advantage of all this data, much of which is dark and inaccessible, and make better decisions from the insights it can produce, while creating net-new or augmenting existing workflows across the entire organization.
Building on projects with thousands of organizations around the world, we’ve observed that organizations choose to:
Speed time to value with pre-built AI apps for common use case (e.g., customer service, financial planning)
Automate knowledge work and business processes
Employ AI-assisted business intelligence and data visualization
Automate planning, budgeting and forecasting analytics
Customize with Industry vertical AI-driven frameworks
Innovate with new business models intelligently powered by AI
Our lead-with products supporting this rung are:
IBM Watson Assistant
IBM Cognos Analytics
IBM Planning Analytics
IBM Watson Discovery
AI is all about the ability to build, deploy, catalog, and manage models, which is what IBM’s AI portfolio provides. It gives you the ability to:
BUILD models for making predictions. Watson Studio supports this.
DEPLOY to put custom models into production, in an application or business process. This is where the model starts to make predictions for the client’s business and it happens with Watson Machine Learning.
CATALOG for data discovery and activation. The Watson Knowledge Catalog allows users to access, curate, categorize, and share data and assets wherever they are.
MANAGE to operationalize and automate the management of models and tools across your business, with trust and transparency. IBM’s answer for that is Watson OpenScale.
All of this together is our AI portfolio that you can access on the private cloud, on the public cloud, on prem, or on desktop.
Watson Application Services are packaged applications that we bundle together to provide the user with AI in a box.
AI Open Source Framework: This is, without a doubt, the future of the market. We built our AI portfolio on open source to make it secure and managed so that our clients can put it to use without the risk or extra cost.
IBM Cloud Private for Data: IBM’s AI portfolio is designed to deploy on our unifying platform. This is our prescriptive approach—one platform, supporting a multi-cloud environment, that brings all of a client’s data and AI capabilities into one set of collaborative workflows and governance capabilities.
Without having to move to a new database we have DS solution that will work for you – coud, local dektop, connectors
“we are not making money on where your data is hosted” – you don’t have to move your data
These are things I’m running into
Building upon previous slides what are we introducing today? Meets your wants while solving your problems
Low to high DS skills spectrum
How does WS fit into the trends – looking at BMC for cloud + WS whatis our value to customer; how are we differentiating from SPSS Modeler?
30 min filling in BMC
30 mins which messages do we want to articulate given strwngths as offering and where we’re going
Goal: crisp & differrentiating
Okay so now diving a level deeper – and again each of these will go back to the 3 values we have on the board
WS = DS platform where your whole team can work together – meaning DS, LOB expert, business analyst, or anyone working on the DS platform – it’s very easy to bring all these people with different levels of expertise on the 1 platform of WS
The reason all these varying levels of expertise can work together is because of these tools you see on the right
DS are rare resources – they each have their own preferences of tools or languages that they want to work with so we have open source tools like R studio and Jupyter Notebooks where DS can come in and write their own code by choosing their own language (R, Scala, whatever they want), and go ahead and get started
At the same time, if someone is not really into coding but has the knowledge, they can use visual tools - like SPSS Modeler, Data Refinery, & Model Builder – where they have a drag and drop UI they can use to build a model normally and get predictions
Both sides are available in the same platform of Watson Studio.
You can give different access to different people.
The whole point is that WS really reduces the barrier between your resources and the DS journey – anyone can get started regardless of their expertise
which brings us back to the strategy of increasing your team productivity
You can see on the right, we provide different technologies and frameworks to support building models - you can use Spark or if you're working with Deep Learning we support all the other favorites so you can choose any of these and work on Studio
Let's say you're into deep learning, but you are not a coder, we have a drag & drop interface where you can build a deep learning model, expose that node, and use it in some other applications
We provide all these cutting edge infrastructure technologies so you can take advantage of any of them while you are working in WS
Again: easy, fast, and increases your team productivity
Again getting started very quickly to increase productivity
So What is Watson Studio Desktop?
What are the key capabilities?
What benefits do you offer?