Uploaded by Ahmed El Agamy

Building an AI-Powered Business: Strategies & Implementation

advertisement
1
LOGIC ExecTalks
Insights for Chairmen, Board Members & CEOs
C O N S U L T I N G
Building The Cognitive Enterprise:
How To Become an AI-Powered Business?
Which regions will gain the most from AI?*
Around 85% of CEOs worldwide
believe that Artificial Intelligence
will significantly change the way
they do business in the next five
years (1).Today, market forces and
consumer demands leave no room
for inefficiencies.
Egypt:
AI contribution of
US$42.7 bn in 2030
(7.7% of GDP)
UAE:
AI contribution of
US$96.0 bn in
2030 (13.6% of GDP)
Saudi Arabia:
AI contribution of
US$135.2 bn in
2030 (12.4% of GDP)
GCC4:
AI contribution of
US$45.9 bn in
2030 (8.2% of GPD)
Source: PwC*
The question now is no longer whether AI will fit into the business, but how to realign the
organization to fully capitalize on it. That is why many organizations, including in Egypt &
Middle East, have rapidly shifted gears to exploit AI. Despite the promise of AI, only 8% of
firms engage in core practices that support widespread AI adoption; they have run only ad
-hoc pilots or are applying AI in just a single business process (2). Indeed, many
organizations still lack the foundational practices to create value from AI at scale. So, How
Can Companies Get Started?
1 Build a Data Ecosystem
To realize the benefits of becoming an
Ai-fueled organization, you will need to have a
data-enabled environment with access to the
right data sets while putting in place more
dynamic data governance, storage, and
architecture. Data could be the biggest
bottleneck as data preparation for training AI
could take up to 80% of the time in any AI
solution development. Thus, the company's
data architecture must be ready and scalable
to support the influx of data that AI initiatives
bring with it. Another consideration is having a
proper data management function to ensure
that data is procured legally, and compliance
standards are met. There are new roles such as
data stewards that help organizations
understand the governance and discipline
required to enable a data-driven culture.
To ensure sustainability of AI transformation projects, leaders need to work on building
a data-driven culture, and this starts by gaining early buy-in of relevant stakeholders as
the quest for AI technology, specifically, can sometimes be puzzling for some internal
stakeholders to grasp. Leaders need to clear up any confusion about the real purpose of
the data, where the data is collected, how it is collected and by whom.
2
2 Identify Strategic AI Use Cases
Strategy is of greatest importance to any AI endeavor, regardless of a company’s
maturity or ambition. But still many companies are following the wrong approach by
making their AI efforts with an IT-centric focus. Leaders must set an AI vision that is
coordinated with the overall business strategy to ensure that AI initiatives get the right
focus, resources, and budget across the organization. Apart from being a technology
for the future, AI is already being used by leading companies to gain a strategic
advantage. You just need to pick the right use case and get started.
3 Create a Roadmap for Yearly Analytics Projects
Any company needs to plan their yearly analytics plan according to the budget, and
this could be done by:
a) Understanding business needs and generating ideas from each objective.
b) Conducting meetings with department managers to determine project value
and complexity (time, cost, and technical requirements)
c) Determining data availability; companies need to define data features, data sources,
format, extraction process and time to integrate the right data insights into the
actual workflow.
d) Drawing a Value-Complexity Matrix to standardize a set of decision-making
business Value
parameters and prioritize projects by evaluating each initiative based on how much
value it will bring vs. its implementation complexity.
High value
Low complexity
Easy wins
Low value
Low complexity
Worth pursuing
or revisiting later
High value
High complexity
Strategic
initiative
High value
High complexity
Deprioritize
Implementation Complexity
4 Choose The Right Domain
The AI domain can be used for solving complex business problems, and for doing this,
it is crucial to have complete knowledge about the latest tools and technologies
associated. Also, business leaders need a clear understanding of which type of AI is
best-suited for their organization’s goals and whether they have the capabilities and
resources needed to support it. So, when choosing a specific domain, consider both
the business and people. There are several AI domains that can create expertise for the
company such as linear regression, classification, neural networks, natural language
processing and computer vision. Some projects can be done through the current
domain the company is working on, other projects will require companies to use a
different domain to drive the needed impact. At the heart of all AI-driven companies
are domain experts who are excellent at execution. So, if you don't have the internal
competencies needed to run a specific domain, you might need to seek external
domain expertise.
3
5 Build an Internal Capacity or Outsource
Once a company has settled on which domain
to work with, many companies face a dilemma
of: should we build or buy? This is rarely an
either-or choice. According to Deloitte’s 2020
State of AI in the Enterprise survey, more than
50% of adopters are buying capabilities instead
of building them, and another 30% are
employing a balanced blend of buying and
building from scratch (3). Yet, organizations
need to go through a period of internal learning
and experimentation before finding out what is
necessary. Companies may decide to buy an AI
solution or build it in-house or even both by
augmenting their existing resources through
investing in AI startups or collaborating with
vendors and other industry partners.
When it comes to talent, the availability of technical talent is one of the biggest
bottlenecks as competition for expertise is increasing with more organizations becoming
AI-fueled. According to McKinsey 2018 Global Survey, respondents say their organizations
are taking an “all of the above” approach: hiring external talent and building capabilities
in-house (4). The core is once you have determined where AI fits into your business
processes, you can evaluate your existing technology, talent, and expertise to determine
where the gaps are and how you could fill them.
Success Factors for Building AI Products
Develop AI Proof of Concept (PoC)
Any business considering AI should undergo a proof of concept (PoC) approach to validate
how well a particular business problem can be solved by AI. The role of the PoC is to enable
decision makers to maximize value and minimize risk through re-scaling the full project
into a pilot to test specific assumptions and validate the viability and feasibility of a
business case, both technically and economically. So, after identifying a business
opportunity that AI could address, get your data in order, bring together the right people,
and develop your PoC. While AI is still in its early days, getting stuck in “pilot purgatory” is
a real risk (5). Nearly, 80% of the AI projects typically do not scale beyond a PoC or lab
environment. Thus, if the metrics show that your PoC is delivering value, companies
should start scaling the technology across the organization and embed it in core business
processes and workflows. A PoC is considered the most cost-effective path as it helps in:
Testing the technologies and methodologies to be used
Delivering more immediate and concrete value
4
Comparing quickly different solutions and approaches
Gaining experience, skills, and confidence in AI
Highlighting impacts on IT infrastructure and the wider business
Identifying and resolving potential data bottlenecks
Set Up an AI Center of Excellence
With scattered data science teams, resources, and legacy systems, it becomes hard to
know where to start. At many businesses, learning typically gets stuck in the mind of one
individual, team, business unit, rather than contributing to the company as a whole. To
make the most of AI, leaders should build a central repository for best practices and
knowledge by having a center of excellence (CoE).
AI has the biggest impact when it is developed by cross-functional teams with a mix of
skills and perspectives; thus, companies need to strategically place their technical teams
within different departments to streamline their business model and nurture agility and
speed. AI CoE is like a core knowledge platform in the company that has all the
accumulative learning from past AI initiatives and a clear vision for use of AI in business
strategy. It brings together the AI talent, knowledge, and resources required to address
challenges of AI adoption, prioritize AI investments, and enable AI-based transformation
project.
Have a Realistic View of AI Capabilities vs. its Limitations
Almost 40% of all
practitioners
who have not yet
invested
in AI do not know
what AI can be
used for in their
business.
Companies need to separate the buzz around AI and think
of it as a technology that will help them solve business cases
efficiently. AI could have a significant impact across
multiple industries and functions, but leaders just need to
think about where exactly the potential for AI is. It is useful
for companies to look at AI through the lens of business
capabilities rather than technologies (6). So, leaders need to
identify the right use cases to leverage AI in their business.
For example, if you are operating in consumer goods, then
demand forecasting is important to you, also that’s where AI
can create the most value. Yet, companies need to be aware
of the big limitations of AI today. Limitations could be
practical such as obtaining large data sets and labelling and
categorizing the underlying data; or technical such as
knowing the “hows” behind the workings of algorithms; or
application limitations such as applying experiences of one
AI model on other use cases.
5
Have a Technical expert/Subject Matter Expert
As stated by The New York Times in 2017, there are fewer
than 10,000 specialists who have enough competence
and expertise to solve serious AI challenges (7). To be an
AI-driven company, you need to have people with both
technical and business acumen to connect the dots
between technology and business impact. But
companies still find difficulty in having AI talent in house
due to current supply-and-demand imbalance, that’s
why external expertise can provide an effective gateway
to bring in AI expertise quickly. Companies need their
own AI experts to develop and customize algorithms,
ensure that they invest in the right AI applications and
services, have access to data sources and draw realistic
AI roadmap given their current levels of talent, data
access, and strategies. Such “Reality Checks” from
technical experts are increasingly important as
companies fundamentally transform the way the
business conducts its day-to-day operations using AI
solutions.
Reality Checks from
technical experts are
increasingly
important as
companies integrate
AI solutions into their
workflows &
processes.
Sources:
1- https://www.pwc.com/gx/en/ceo-survey/2019/report/pwc-22nd-annual-global-ceo-survey.pdf
2- https://hbr.org/201907//building-the-ai-powered-organization
3- https://www2.deloitte.com/content/dam/Deloitte/us/Documents/deloitte-analytics/aireadyaiwhat youneed-to-jump-start-and-scale-ai.pdf
4- https://www.mckinsey.com/featured-insights/artificial-intelligence/ai-adoptionadvancesbutfoundational-barriers-remain
5- https://www.mckinsey.com/featured-insights/artificial-intelligence/ai-adoptionadvancesbutfoundational barriers-remain
6- https://www2.deloitte.com/eg/en/pages/financial-services/articles/artificialintelligenceinsurance
industry.html
7- https://www.nytimes.com/201722/10//technology/artificial-intelligence-expertssalaries.html
Download