Scribd
Upload
3 views

Introduction to AI & ML - Technology & Data

The document outlines the essential components for successful AI technology initiatives, emphasizing the importance of clear business objectives, a broad AI toolkit, and effective data management. It discusses the need for high-performance computing environments, the challenges posed by legacy systems, and the significance of automated machine learning in speeding up model development and deployment. Additionally, it highlights the rapid pace of change in AI technology, necessitating continuous adaptation and monitoring by organizations.

Uploaded by

Rahul Marshall
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
3 views

Introduction to AI & ML - Technology & Data

The document outlines the essential components for successful AI technology initiatives, emphasizing the importance of clear business objectives, a broad AI toolkit, and effective data management. It discusses the need for high-performance computing environments, the challenges posed by legacy systems, and the significance of automated machine learning in speeding up model development and deployment. Additionally, it highlights the rapid pace of change in AI technology, necessitating continuous adaptation and monitoring by organizations.

Uploaded by

Rahul Marshall
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 14

TECHNOLOGY & DATA

• Clear Business objectives for AI technology initiatives


• Supporting a broad range of AI use cases by creating a broad AI
toolkit
• Building applications faster and better with tools like automated
machine learning
• Achieving a broad scale of AI deployment
• Managing and improving data for model training and other
purposes
• Dealing with legacy applications and complex technology
architectures
• Building or sourcing a high-performance computing infrastructure
for AI
• Improving IT operations with AI
• Using All the Tools in the Toolkit
• Companies that compete on AI realize that there are many different
AI technologies, and they are usually willing to make use of all of
them.
• Different technologies are useful for different use cases, and
organizations that adopt AI broadly and deeply have breadth in
their use cases and the technologies that are applied to them.
• Building AI Applications Faster and Better
• Our organization has embraced AI as key to its future, probably like
things to move a bit faster.
• Specifically, new AI algorithms to be developed more quickly by
more data scientists.
• Solution engineering
• After the necessary training data has been gathered, most machine
learning efforts in companies are focused on the development of
models, but 84% was interested in a broader focus.
• In the past, solution engineering would require rethinking the
problem to stay within resource constraints.
• Automated machine learning technology can substantially lessen
those resource constraints.
• Solution engineering is still necessary, but the horizon of solutions
has broadened.
• Model development
• In the model development phase of the methodology, data is
analyzed, variables or features are engineered, and the model that
best fits the training data is identified.
• AutoML using DataRobot speeds this phase of the process
considerably, increasing the productivity of data scientists.
• That frees them up to fit more models and/or to give more effort to
other high-value aspects of the process (e.g., solution engineering,
feature engineering, and so forth).
• Model deployment
• The third and final component of the 84.51° approach to
machine learning is model deployment, in which the
chosen model is deployed in production systems and
processes
• Automated machine learning tools can help with the
deployment process by generating code or APIs that
embed the model.
• Getting to Scale
• One of the key challenges for many organizations with AI is getting
to a sufficient scale to make a difference in their operations and
performance.
• Technology can help companies achieve this objective, though as
with the other AI objectives the complete answer is to combine
technology with other changes like new Processes and newly
involved groups of people.
• Managing Data for Training and Everything Else
• Data is the precursor of machine learning success, and models
can’t achieve accurate predictions without large quantities of good
data.
• Every organization that’s serious about AI must deal with its data at
some point—structuring or rearchitecting it, putting it on a common
platform, and addressing pesky issues like data quality, duplicated
data, and siloed data (repository of data) throughout the company.
• It’s fair to say that the single biggest obstacle for most
organizations in scaling AI systems is acquiring, cleaning, and
integrating the right data.
• Data environments for AI-oriented companies have several
characteristics:
• Most are cloud-based
• The data they use is machine readable
• They involve internal and external data
• They are centralized
• They have a different focus
• They use new systems
• They are adding team members
• The Burden of Legacy Applications and Architectures
and How to Deal with Them
• Many companies have old, fragmented legacy systems, which
makes integration challenging.
• In many cases, they need to modernize these systems in order to
integrate AI capabilities with them.
• Companies in this situation have to get control of these diverse
architectures and simplify them over time.
• AI, Digital, and AIOps
• One of the most popular applications for AI in recent years,
according to Deloitte’s annual survey of AI activity, is IT itself.
• AI and automation capabilities can predict and diagnose problems
in networks and servers, and automation programs can restore
them to health.
• The use of AI to help with IT operations has been called IT
autonomics, but more recently goes by the term AIOps
• AIOps involves software and IT device data to identify problem
areas and automate aspects of IT operations.
• Building High-Performance Computing Environments
• Companies that plan to do significant amounts of AI
development need to establish a suitable hardware
environment. Often called high-performance computing or HPC
environments, they typically include systems that can perform
very fast numeric calculations in parallel.
• Deep learning-based AI models use graphics processing units
(GPUs) that are available in both cloud and on-premise
configurations
• Companies also need considerable storage for the extensive
data required to train machine learning models, and may
require low-latency architectures for real-time scoring of
models.
• Other types of AI methods require powerful versions of regular
processors.
• The Pace of Change in AI Technology
• AI technology is perhaps the fastest changing of any information
technology domain.
• No organization should expect that they can establish a technology
environment for AI and let it ride for a decade.
• Constant monitoring of external offerings and their match (or
mismatch) to internal needs is critical in the AI space.

You might also like