Ai Potential 8 Steps To Success
Ai Potential 8 Steps To Success
Ai Potential 8 Steps To Success
into AI results:
8 steps to success
Your plan for unleashing the power of
AI everywhere. That’s AI at work.
Authored by:
Cam Lau, Eugene Wyniawskyj
Strategic Engagement and Transformation, UiPath
Contents
3 Introduction
4 Unleashing AI potential though automation
5 Defining mission objectives for AI success
6 Building a transformation action plan
6 - Step 1: Understand the business’ strategic problem statements
8 - Step 2: Identify the top-down value at stake
9 - Step 3: Verify the inside-out value potential
10 - Step 4: Define the software and data reference architecture
12 - Step 5: Envision the future of AI-powered automation
13 - Step 6: Scale up and down the value chain
15 - Step 7: Operationalize AI
16 - Step 8: Build a holistic investment case on total cost of
ownership and AI potential
17 Achieving strategic outcomes
However, for many C-suite executives, the personal ambition • Low ROI realized and diminishing value
to lead on AI and automation transformation is a constant delivered to date (e.g., productivity becomes
battle of skill, not will. Historically, AI and automation a key metric, not profit and loss (P&L)
transformations have been complex, highly federated, legacy impact, etc.)
beasts—often starting from the subset of macros and API • Value not sized in logical units a business
integration to more avant-garde initiatives of data science unit and/or function can actualize (e.g.,
and new application business building. However, the most there’s no value in automating a subprocess
common iteration of AI and automation programs are ‘digital’ that happens for five mins—once a year)
variants of Lean Six Sigma-style operational excellence
• Automation programs stuck in piloting or
programs.
PoC scope
UiPath is familiar with common objections and challenges • Prohibitive cost of implementation and
when embarking on this kind of program. The most common maintenance due to the increasing costs of
one is “What’s the business case and value of AI and niche tech talent
automation?”. • Avoiding technical debt across the complex
ecosystem of different point solutions
The best place to start is to look at the strategic objectives that may be in flight in the organization, this isn’t a
set by the board. It’s important to understand the problem comprehensive list. It does, however, offer a useful context
statements and corresponding objectives before embarking for the kind of transformation planning needed today.
on a transformative change journey. When asked about
problem solving, Albert Einstein famously said, “If I had an Defining enterprise problem statements that can be
hour to solve a problem—I’d spend 55 minutes thinking about addressed with AI isn’t easy. But, by looking at what
the problem and 5 minutes thinking about solutions.” drives value for your business, it’s possible to focus on
the concepts and principles that make up the rest of the
Figure 1 illustrates different types of mission goals transformation plan.
across enterprise, divisional, and process transformation
archetypes. Considering all the different transformations
FIGURE 1
Foster enterprise agility by Digitize the division across Reduce waste in time, effort,
nurturing creativity, collaboration, non-digital products, services, or and materials to drive operating
and innovation operations expense efficiency
A platform for
A gateway to latest intelligent and New or upgrade: processes all stages of Improve
cognitive tech and technology improvement
FIGURE 2
Innovation Resistance
Sizing up what will solve the problem statements is key to These can include operating margin, profit and loss (P&L), days
justifying the reasons for embarking on the execution and sales outstanding (DSO), working capital, net promoter score
delivery of these levers. A proper value analysis helps set (NPS) /customer satisfaction score (CSAT), and even risk and
up a successful transformation by sizing the opportunity compliance. These value drivers end up being hugely important
along with the most important enterprise value drivers. KPIs to measure success.
FIGURE 3
Better business outcomes, teams work Good start but not scalable and holistic
together to achieve common goals
Identifying the value potential can be an extremely difficult approach, as shown in Figure 3. UiPath has worked with
and manual task. That’s especially true when there’s customers that have faced a wide range of scenarios, and
no process or operational taxonomy, the data quality this is where technology can really help. They’ve leveraged
in telemetry is poor, or there’s a lack of business and top-down mining technologies like UiPath Communications
operational teams in place to track and gather insights. Mining and UiPath Process Mining, in combination with
proprietary AI and automation benchmarks. This meant
Defining the value of aligning your program to core value they could build a data-driven fact base that allows them
drivers and problem statements helps the transformation to identify (without the need for deep SMEs) the “big rock”
to focus in on areas of specific attention. Most importantly, opportunities for digitization, LEAN, automation, and AI
it identifies the specific AI and automation levers that are within an entire business. Furthermore, it helped them
needed to be deployed. set up the foundation to measure the effectiveness of all
transformation initiatives in-scope.
A top-down approach drives better outcomes when
compared to the more commonly seen bottom-up
A top-down sizing of value potential needs bottom- considered for applicability and effectiveness
up verification. Historically, this part of the strategic at this stage.
transformation is delivered through deep-dive interviews and
workshops run by an ecosystem of consulting partners. Here, analysis often incorporates existing automation
maturity and AI penetration to shed light on the factors that
The objective of this step is to capture key process information erode the aspirational automation potential. A platform-
and pain points, and to develop a target-state architecture in centric approach at this stage is usually articulated with
the business. Specific product features, off-the- shelf resources great detail and benefits extend beyond the hard savings of
like solution accelerators and marketplace use cases, are specific use cases alone.
FIGURE 4
40%
Increased success
implementation cost chance
200%
Scale AI and automation libraries Faster time to
value delivery
20%
Reduce IT OPEX TCO Reduction in
OPEX services
Source: 2023 - Strategic Engagement Study of High Maturity Fortune 500 Customer
Leading customers leverage UiPath and its consulting However, real-world use cases are only a fraction of the total
partners to bring to life proprietary industry and line value at stake. Figure 4 illustrates how P&L impact beyond
of business data sets, combined with task and activity the hard savings can be achieved by embracing an overall
capturing technologies—like UiPath Task Mining—to align to AI-powered automation platform.
the daily practicalities of business operations. This results in
a confident and actionable pipeline of use cases.
Once we know where to play, the next step is making sure data integrations. Similarly, a strategy that just focuses
we know how to play. It’s important to thoroughly analyze on core RPA—without exploring how digital assistants
the enterprise technology stack. Only by identifying the and Gen AI fit into the workflow—won’t maximize the
potential blockers here can enterprise risk and agility can value the technology could bring to the operation. Finally,
be managed and optimized. clients often have myriad intelligent automation tools
in their enterprise stack, and miss the value potential of
Defining what ‘good’ looks like across the enterprise consolidating and reducing the number of point-solution
architecture’s core systems of record, differentiation, and vendors.
innovation—overlayed with AI and automation use cases
that span across the stack—is key to knowing how to Figure 5 illustrates how a system of AI-powered
leverage these technologies strategically. automation should be treated as omnipresent—
surrounding and running throughout the enterprise’s
For example, a strategy that solely relies on API integration core systems.
to connect systems of record together will fall short in
capturing the full potential brought about by user interface
FIGURE 5
Systems of innovation
Consumer focused, new applications, business models
Continuous discovery and monitoring
LLMs and AI
Low-code apps Workflows Analytics
expert systems
UI and API integration
Systems of differentiation
Unique processes and capabilities, SaaS, modules of a platform
Systems of record
Core transactional processing mission critical master data
Source: 2023 - Strategic Engagement Study of High Maturity Fortune 500 Customer
Imagine a world with 10X more creativity and inspiration, example of how the activities and processes of a typical back
starting new businesses and entering new markets at office worker may be augmented and streamlined with AI and
a moment’s notice, where every employee is 10X more automation in the loop.
productive. Deployed correctly, AI-powered automation
creates an environment in which your employees thrive. AI can impact the way an organization manages people and
culture, because lasting change and sustainable business
Another way to envision the benefits is through the lens impact require a holistic uptake.
of your employees’ daily activities. Figure 6 provides an
FIGURE 6
Human-in-the-loop 1 2 3 4 5 6 7 8 9 10
process automation
Individual general
knowledge tasks
(automation + Gen AI)
Back-office
Physical activity employee
Human-in- UI + API
Automation capabilities UI + API integrations
the-loop + integrations
Digital assistants + Generative AI models
often deployed
Cultural adoption is a key workstream for any transformation Making sure that executives are aware of your AI and
plan. The plan may also align with broader corporate change automation ambition—and how people, processes, and
strategies and in-flight business programs. This could include technology stack will be impacted—will deliver faster and
tapping into existing capabilities and change management more scalable programs.
content being delivered as part of an SAP S/4 Hana migration,
Agile customer journey re-design, or business process
optimization (BPO) offshoring migration.
Guaranteeing that any AI-powered transformation program (E2E) is often overused, customers that achieve it follow a
can achieve its full potential requires a perspective on how two-speed approach, as shown in Figure 7:
it can scale across the entire value chain. While end-to-end
FIGURE 7
Long-term
Design thinking, new process design, and process redesign
EBITDA impact
The first is strategic. The full end-to-end redesign of a process tactical approach. Often, the strategic horizon leans on large
or operation using digital technology, AI, and automation programs of product engineering, new business building and,
is the primary objective and end-state. Typically, this sees in extreme cases, mergers and acquisitions (M&A) to actualize
the reimagination of entire cross-functional processes and the value over multiple years.
customer journeys.
However, this approach isn’t without its risks and challenges.
The strategic approach takes the best of design thinking That’s why a comprehensive strategy always includes a plan
practices and overlays an end user appreciation for the for a second speed. This rapid and tactical speed is pragmatic
target state journey. This drives more holistic business and underpins the in-year business case through tactical line
outcomes, often on an order of magnitude higher than a more of business delivery.
Operationalizing and scaling AI requires a holistic operating Our leading customers set up holistic operating models to
model that’s agile, comprehensive, and proactively resilient. capture ‘who’ is needed and ‘how’ AI can be scaled across
Mapping the interactions across an enterprise in this the enterprise.
operating model helps manage risk—identifying potential
issues before they become problems. This strategy is anchored on the following four principles:
1. Being realistic with existing automation and AI maturity: regardless of whether an enterprise is starting from a
greenfield or brownfield environment, a large upfront focus on the operating model design will depend on existing
support processes, as well as discover and delivery capabilities. Determining whether existing processes are enough
or new ones need to be designed.
2. Tightly linking organizational priorities, OKRs, and KPIs: translating and cascading corporate strategic KPIs
to the individuals in the business unit is key to ensure success and continuous adoption of any AI and automation
program. Understand how each objective/KPI can be impacted by the different components of the platform
technologies at your disposal.
3. Maintain a focus on agility, but don’t forget about the merits of waterfalling: agility is key to delivering
transformative impact with AI-powered automation. Because of the tight dependency on business operations, Agile
delivery methodology in isolation won’t work. Deploying AI-powered automation requires a real focus on change
management to mitigate the associated disruption to live operational processes, teams, and customers. Not just the
use-case definition and specific process improvements it’ll bring.
4. Achieving competitive advantage without jeopardizing ethics: as AI finds its way into every aspect of the
business, strict guidelines must be in place to ensure democracy, legitimacy, and transparency as AI and automation
are scaled and operationalized. Defining the personas that need the most attention and actively managing AI
training data inputs to minimize unforeseen biases is the best first step to ensure ethical safeguards. For example,
business processes and technological guardrails will prevent AI-driven processes from discriminating against end-
consumers. These guardrails range from human-in-the-loop reviews, to strict data cleansing and model retraining
ML operations. A plan that captures the widest range of scenarios is key to making sure AI and automation are
operationalized sustainably and successfully.
Leading UiPath customers adopt a hybrid approach to their delivery and enablement. Plus, a highly iterative top-down
operating models, built on the understanding that AI systems diagnosis, discovery, and delivery focuses the team on value
aren’t infallible and require human supervision. This hybrid capture, ethical deployment, and security.
model involves agile methodology and scrum teams driving
Formulating an investment case that forecasts the real value AI is at this same inflection point, akin to the early days of
of an AI program can only be done by including total cost RPA, where point solutions and vendors vied for the same
of ownership (TCO) in your analysis. Accurately calculating fixed budgets. Should history be repeated, the inefficiencies
the impact of new roles and people, new technology and of managing an estate of disparate technologies will (again)
infrastructure, and new investments in change management drive up IT operational costs and increase TCO.
is needed to get to a real ROI figure.
In contrast, a unified platform-based approach like UiPath—
Our customers often model out benefit growth due to near- especially with its focus on Gen AI and Specialized AI
term emerging and improving technologies, while conversely integration for the whole enterprise—will do the reverse.
becoming more aggressive when modeling out costs due to
economies of scale and emerging unified platforms.
It’s important to note that Gen AI and Specialized AI are
much like RPA back in 2016. When these technologies were
introduced into the enterprise in isolation, companies faced
exploding IT operating costs and shadow IT. It wasn’t until
the consolidation of the market and a holistic platform-based
approach started to emerge that ROI started to turn positive.
1. Cost transformation: cost transformation means achieving better results, having spent time and money
wisely. Introducing AI-powered automation into the workplace allows businesses to streamline more
processes, reducing the need for manual interventions and reorienting employees to high-value work. The
result is more efficient and agile processes, with a greater focus on customers and business outcomes.
2. Profitable growth: increasing profits while maintaining or even reducing margins across your business. By
introducing AI-powered automation into the workplace, organizations can improve customer engagement,
increase productive innovation, and launch new services and products more quickly. The result is higher
margin growth and expanding revenue streams thanks to a renewed focus on customer acquisition, retention,
and growth.
3. Managed risk and maximized compliance: a combined AI and automation approach can play a crucial role in
managing risk and compliance challenges. By implementing an intelligent platform that quickly incorporates
new compliance rules, enterprises can keep up with evolving regulations and reduce the risk of non-
compliance. The business impact is significant, driving efficiency, reducing costs, and increasing customer
trust.
4. Improved customer experience and increased employee engagement: by streamlining systems and
processes, organizations can enhance the overall customer and employee experience. Businesses that use
automation to improve employee experience can attract and retain the best talent. They’ll not just become
an employer of choice, a happier workforce will boost customer satisfaction through improved interaction,
support, and service. Which will drive loyalty and revenue growth.
To see customer case studies and learn more about how to unleash
the potential value of AI-powered automation, visit uipath.ai.
UiPath is AI at work.