Is your AI and ML strategy too siloed?
By Mike Salvino, Andersen Alumnus and currently DXC Chairman, President and CEO
Copyright
2023 DXC Technology. This article originally appeared in DXC’s Insights -Our
Perspectives, original copy can be found
here.
Reprinted with permission. No further reproduction is permitted without
permission from DXC.
Four
years ago, I wrote a similar article, and now, as the Chairman, President and
CEO of DXC Technology, I am reminded of its ongoing relevance as I read the
latest headlines and engage with customers worldwide who are eager to assure
their boards and investors that they are embracing innovation to drive growth.
Every
CIO is looking to invest in the latest artificial intelligence (AI) and machine
learning (ML) technologies to gain a competitive advantage, achieve cost
efficiencies, accelerate speed to market and enhance the overall experience. Believe
me, you don’t want to be late to this party. But the pressure to act fast may
hinder your ability to achieve desired business objectives.
One
aspect that many CIOs overlook as they embark on their AI/ML journey is how to
evolve their talent strategy to effectively scale their initiatives in this
field.
The
new 80/20 rule
At
DXC, we have learned that to maximize the benefits of machine learning for your
business, you need to adopt a hybrid approach that combines technology to take
you 80% of the way there, and people who will take it the rest of the way.
Many
software products today boast embedded AI and ML capabilities, suggesting that
businesses can simply rely on a plug-and-play approach. Take it out of the box,
plug it in and it magically works.
But
the true essence of machine learning — its groundbreaking nature and the value
it brings to businesses — is impossible to deliver without human intervention.
A
perfect example of this is DXC Platform XTM, our data-driven intelligence
platform that efficiently manages our customers' IT estates. Platform X
integrates advanced AI and ML technologies with a vast catalog of automation
bots. However, it is designed to include our engineers, who operate from a
virtual "control room" to ensure its optimal performance.
How
ML works
To
make machine learning effective, you must first feed your data into the
platform, clean it, configure the ML model and then calibrate the model as it
encounters data. This process of running data through the model, and
continuously monitoring and gathering feedback, improves the model's
performance and generates accurate results.
To
put it in perspective, consider the human brain. People are better at thinking
and drawing conclusions when they have more data, experience and accumulated
wisdom to process that data.
Relying
solely on what's written on the ML "product box label" or the
vendor's response to an RFP can lead you down a risky path. Every CIO should
prioritize three critical success factors when implementing ML in their company.
1. An ML model without data is like a car without gas.
To train machine learning models effectively, you need
high-quality and high-volume data. This data must be easily accessible, in the
right format, and diverse enough to ensure unbiased results.
At DXC, our data model is built on more than 60 years
of managing essential systems for over 6,000 customers. This rich history
enables us to train our models with better quality data, resulting in faster
and more accurate recognition of service-impacting issues, leading to fewer
disruptions. Each individual customer benefits not only from their own data but
also from the collective wisdom derived from our customer base. In other words,
our data (secured and anonymized) becomes your data.
2. There is no "set-it and forget-it"
solution.
As mentioned earlier, ML models require human
oversight. While software solutions will continue to improve over time, the
success of ML in business relies on skilled professionals who can make it work
effectively.
At DXC, we have an elite team of specialized data
scientists who experiment, design and create models. Our global engineering
workforce across 70 countries enables us to deploy, monitor, audit and optimize
models at scale.
One crucial capability our engineers possess is the
ability to determine the right mix of ML products and features. Many products
offer similar functionalities, but the knowledge of what to apply where and
when for optimal results is a uniquely human trait that requires practical
experience and judgment.
While AI and ML may disrupt certain job markets, they
also create new and exciting opportunities for tech-savvy individuals who
embrace change.
3. Don't forget about integration.
An often-overlooked aspect of the plug-and-play
approach is the effort required to configure all the integrations. Only new
startups have the luxury of building a greenfield estate. The majority of us
are faced with the challenge of integrating new technologies with existing
ones. Most Fortune 500 companies have sprawling and complex IT estates,
comprising thousands of servers, hundreds of strategic applications, and
distributed endpoint devices supporting a dynamic and often virtual workforce.
This
is where DXC Platform X excels. Its open, modular architecture allows for easy
integration and flexible deployment options, leveraging our customers' current
and future IT investments. We have done the heavy lifting by providing
pre-configured integrations for top enterprise SaaS platforms, ready-to-use ML
models, a catalog of hyper-automation assets and experienced engineers who
seamlessly connect all the pieces.
Summary
The
AI and ML revolution will undoubtedly continue to accelerate, driving
advancements across all industries. At DXC, we are excited to be at the
forefront, thoughtfully evaluating and applying emerging technologies for the
benefit of our customers and our people.
Regardless
of your strategy — whether you choose to buy, build, or partner for AI and ML
advancements — you must prioritize more than just ML itself. High-quality data,
the right talent, and a well-rounded integration strategy.
About Mike Salvino He is chairman, president and chief executive officer of DXC Technology, the world's leading independent, end-to-end IT services company, with over 130,000 colleagues serving nearly 6,000 private and public-sector customers in some 70 countries from a diverse array of industries.