Build a Scalable AI Deployment Plan

Your implementation and scaling plan is about more than just technology.

Analyst Perspective

Your implementation and scaling plan is about more than just technology!

Aggressively scaling artificial intelligence (AI) with the latest concepts and leading-edge technologies is an attractive ambition for many organizations. However, rolling out AI as quickly and as broadly as possible often comes with costs that outweigh the benefits and may introduce cultural, security, and business operational risks and changes for which the stakeholders have no appetite. The lagging development of a good AI practice can further hamper the future returns of today's investments.

Start small with the intent to learn and scale. The right AI initiative helps you learn the new AI technologies, their measurable impacts, and how they benefit your teams while allowing your team to establish good foundational capabilities and build the necessary relationships and collaborations for you to be successful. These factors will then allow you to explore more sophisticated, complicated, and innovative opportunities to drive new value to your team, department, and organization.

Andrew Kum-Seun
Research Director,
Application Delivery & Management
Info-Tech Research Group

Bhavya Vora
Research Analyst,
Special Projects
Info-Tech Research Group

Executive Summary

Your Challenge

• AI adoption has gained significant momentum as your business leaders see the positive outcomes in your proofs-of-concept and pilot projects, such as improved operational efficiencies, cost optimizations, and higher output quality.
• Your stakeholders are ready to increase their investments in more AI solutions. They want to scale their initial success to other business and IT functions and technologies.
• An implementation plan is needed to ensure the right projects are defined and prioritized to meet the AI initiative goals and the required execution capabilities are in place to support AI growth.

Common Obstacles

• The shift from small-sized AI adoption to complex, broad AI implementations brings new challenges and barriers to your organization:
• Stakeholders are oversold on the expectations of AI and propose aggressive initiatives that are costly, risky, or impractical.
• There is little motivation or tolerance to change existing business operations or enterprise systems to see the full value of AI.
• Teams are ill equipped to meet the demands and complexities of scaled AI implementations.

Info-Tech's Approach

• Build a scalable deployment plan for your AI initiatives by following Info-Tech's guided approach:
• Identify and build an implementation plan, including a roadmap and high-level backlog, for a single high‑priority initiative.
• Identify the gaps in your ability to execute on an AI‑focused plan.
• Identify and prioritize your AI scaling initiatives to support your deployment of AI at scale.
• For an example initiative, break down the cost of implementation and maintenance.
• Repeat the practices in this research for other AI initiatives.

Blueprint Insight Summary

Overarching Info-Tech Insight

Position your AI deployment plan as a collectively owned and managed artifact focused on sharing, enabling, and continuously improving. AI requires various roles, capabilities, and technologies to harmoniously work together to gain the outcomes your stakeholders expect to see. Everyone needs to participate in the planning to ensure all perspectives are accommodated in the solution design and the right support is in place.

Phase 1

Treat AI like any other enterprise technology. Implementing AI involves understanding more than just how it impacts the users, customers, and bottom line. You need to know how these improvements meet risk and change appetites and tolerances, how AI impacts enterprise data and systems, and how outputs drive both financial and non-financial business priorities.

Phase 2

Mature your practice as you scale your AI technologies. Good skills, resources, and governance and management practices are critical for the successful scaling of AI tools and technologies. AI is not a set-it-and-forget-it asset; it needs to be delivered with quality in mind and continuously monitored, reviewed, and maintained.

Phase 3

Implement AI with the future in mind. AI motivates the implementation, modernization, and enhancement of key tools and technologies to maximize the return on those investments. The resulting technology governance and management practices act as building blocks in your organization. We must expect that more AI and other enabling technologies will be implemented using earlier investments.

Phase 4

View AI costs as organic. AI is designed to continuously consume information and resources so that it can grow and improve over time. It has a lot of moving parts (e.g. AI model, data integrations), each of which requires specific care and feeding, complicating set up and management.

Info-Tech offers various levels of support to best suit your needs

DIY Toolkit

“Our team has already made this critical project a priority, and we have the time and capability, but some guidance along the way would be helpful.”

Guided Implementation

“Our team knows that we need to fix a process, but we need assistance to determine where to focus. Some check-ins along the way would help keep us on track.”

Workshop

“We need to hit the ground running and get this project kicked off immediately. Our team has the ability to take this over once we get a framework and strategy in place.”

Consulting

“Our team does not have the time or the knowledge to take this project on. We need assistance through the entirety of this project.”

Diagnostics and consistent frameworks used throughout all four options

Info-Tech's methodology to Build a Scalable AI Deployment Plan

Note: The goal of this research is to complete a scalable deployment plan for one AI initiative with the intent that the tools, tactics, and techniques be repeated for other initiatives.

Your deliverables

Leverage these supporting deliverables to help you accomplish your goals.

AI Deployment Plan Report Template

Finalize a report for your leadership on what your AI implementation and scaling plan can look like.

AI Execution Capability Assessment Tool

This tool will help you analyze your organization's foundational AI capabilities and identify the areas that most need attention.

Workshop Schedule: Build a Scalable AI Deployment Plan

Note: The goal of this workshop is to complete a scalable deployment plan for one AI initiative with the intent that the tools, tactics, and techniques be repeated for other initiatives.

0.1 What do we want to accomplish?

1 hour

1. Write what you want to accomplish by maturing and scaling your AI capabilities using a practice user story.

Input

• N/A

Output

• Desired outcome of expanding AI adoption

Materials

• Whiteboard/flip charts

Participants

• CIO and direct reports
• AI leaders and teams
• Business representatives related to the AI initiative

Deploy AI at scale

Deploying at scale requires careful consideration of several different elements.

1. AI Solution Design

Deconstructing the AI initiative into granular steps – building use case models, charting system, data, and process flows – simplifies the initiative, lays the groundwork for clarity, and accelerates AI integration across business functions.

2. AI Capabilities

Establishing a frugal AI mindset helps prioritize actionable progress over perfection. Concentrate on leveraging existing strengths and making strategic improvements to gaps to scale AI.

3. Roadmap and Backlog

Crafting a roadmap and maintaining a prioritized backlog of AI initiatives serve as a strategic compass, guiding phased deployment of AI. Estimate and prioritize the backlog of initiatives, assessing degree of effort and value.

4. AI Budget

Quantify the impact of scaling the initiative by conducting T-shirt sizing estimations to evaluate cost factors against the backdrop of risk and value potential.

Info-Tech Insight

Position your AI deployment plan as a collectively owned and managed artifact focused on sharing, enabling, and continuously improving. AI requires various roles, capabilities, and technologies to harmoniously work together to gain the outcomes your stakeholders expect to see. Everyone needs to participate in the planning to ensure all perspectives are accommodated in the solution design and the right support is in place.

Phase 1

Build an Actionable Implementation Plan

Build on your early AI success

Notable initial benefits:

QUALITY

Create high-value and high-quality outputs consistently, and proactively identify and address issues before they occur.

TIME SAVED

Reduce the time spent by employees on completing manual, repetitive, and error-prone work.

CUSTOMER EXPERIENCE

Improve the customer experience with a product or service via reliability, engagement, transparency, etc.

PRODUCT GROWTH

Expand the key products and services to ultimately drive revenue expansion or customer impact.

Can we gain more from AI?

As industries evolve and adopt more tools and technologies, their business operating models become more complex. Task-based and desktop-based AI are often not enough. More sophisticated and scaled AI is needed to simplify and streamline complex operations from end to end and align them with organizational goals.

1.1 Conduct a retrospective on previous AI initiatives

1 hour

1. Discuss the lessons learned from your previous AI initiatives, pilots, and experiences:
   • Start – What should we start doing for future AI initiatives?
   • Stop – What should we stop doing in future AI initiatives?
   • Continue – What should we continue doing in future AI initiatives?
2. Document the results of this exercise in the AI Deployment Plan Report Template.

Download the AI Deployment Plan Report Template

Input

• Experiences from past AI projects and implementations

Output

• Lessons learned from past AI projects and implementations

Materials

• Whiteboard/flip charts
• AI Deployment Plan Report Template

Participants

• CIO and direct reports
• AI leaders and teams
• Business representatives related to the AI initiative

Why do we want to broadly leverage AI?

The business needs AI to be successful

"94% of business leaders surveyed agree that AI is critical to success over the next five years" (Deloitte, 2022; N=2,630).

Maintain the momentum from early successes

Lowering costs, discovering valuable insights, and improving collaboration across business functions or organizations were the top three AI outcomes that were achieved to a high degree (Deloitte, 2022, N=2,630).

AI is becoming the way of working

"Seventy-nine percent of all respondents say they've had at least some exposure to [generative] AI, either for work or outside of work, and 22 percent say they are regularly using it in their own work" (N=1,684, McKinsey, 2023).

Peers and competitors are investing in AI

"In fiscal year 2022, U.S. government agencies allocated $1.7 billion to AI R&D, up 13% from the year prior and an increase of 209% from 2018" (Stanford Institute, 2023).

Enterprise AI adoption requires significant top-to-bottom investments

Expanding AI adoptions requires organization-wide investments, such as:

• A unified, open analytics platform that can handle large and complex data, support various data processing and analysis models, and enable fast and reliable deployment and management of AI solutions.
• A flexible, cost-efficient, and scalable data management architecture (e.g. data lakehouse, which combines the best features of data warehouses and data lakes), to provide a single source of truth for data, analytics, and AI.
• Privacy-preserving computing techniques, such as fully homomorphic encryption and differential privacy, to protect sensitive data while allowing computation and collaboration across organizations.
• A practice that has the competencies, influence, and empowerment to address dependencies and conflicts across the organization's functional groups.

1.2 Discuss the future role of AI

0.5 hours

1. Discuss how you envision the role of AI has evolved since your earlier pilots and proofs-of-concept. Ask yourself the following questions:
   • How can AI complement and augment employees and externally facing products and services?
   • What are the broad outcomes we want from AI?
   • What is the familiarity with and reception of AI across the organization?
   • What are your peers and competitors doing with AI?
   • What are the policies, restrictions, and regulations that are constraining AI adoption and implementation?
2. Document the results of this exercise in the AI Deployment Plan Report Template.

Download the AI Deployment Plan Report Template

Input

• Lessons learned from past AI projects and implementations

Output

• The organization's perspective on the future role of AI

Materials

• Whiteboard/flip charts
• AI Deployment Plan Report Template

Participants

• CIO and direct reports
• AI leaders and teams
• Business representatives related to the AI initiative

Define an effective AI deployment plan using your AI strategy

1. AI Mission & Vision

Statements that describe the future achievements and promise of AI and articulate what IT does to achieve the vision.

2. AI Principles

Core values, thinking, and boundaries that guide the analysis, perspectives, decisions, and approaches of AI and the target-state design.

3. Strategic Goals and Priorities

Specific corporate financial and non-financial priorities that the organization aims to achieve.

4. AI Objectives and Metrics

Practical, achievable, and measurable expectations of AI initiatives.

See our AI Strategy Presentation Template to get started in building your AI strategy

Establish a repeatable planning process by focusing on one AI initiative

See our AI Strategy Presentation Template to build your backlog of AI initiatives

Gain these key lessons from your first AI initiative:

Verification & Validation

Test the fit of AI capabilities, organizational acceptance, and IT's readiness to support AI.

Foundations

Establish a foundational practice to support future scaling and accommodate continuous learning.

Justification

Rationalize the return on your AI investments and make the case for further AI adoption and innovation.

Prioritize the initiatives that support your business goals

Example:

1.3 Select your initiative

0.5 hours

1. Review your backlog of prioritized AI initiatives. If available, refer to your AI strategy document.
2. Select the initiative to begin your AI journey. Consider the following in your selection:
   • Feasibility – Do you currently have the capabilities to deliver on this opportunity? Do you have the right partners, resources, or technology?
   • Desirability – Is this opportunity a change the stakeholder needs? Does it solve a known pain point?
   • Viability – Does this initiative have an impact on the business value drivers of the organization? Is it a profitable and productive opportunity that will support the business model? Will this opportunity require a complex cost structure?
3. Document the results of this exercise in the AI Deployment Plan Report Template.

Download the AI Deployment Plan Report Template

Input

• Backlog of AI initiatives
• AI strategy

Output

• Selection of AI initiative to be the focus of this plan

Materials

• Whiteboard/flip charts
• AI Deployment Plan Report Template

Participants

• CIO and direct reports
• AI leaders and teams
• Business representatives related to the AI initiative

Reveal the risk and challenges of your initiative through multiple perspectives

Decompose your AI initiatives into smaller deliverable increments. This activity reveals the underlying complexity of the initiative and gives greater confidence in what can be reasonably delivered against defined milestones or timelines.

Use Case Models

Understand the interactions between the system and human users. This helps understand how AI can interact with related systems.

Business Process Flows

Describe the series of steps a person follows to generate the desired outcome. Here, we see how AI changes how employees operate.

System and Data Flow Diagrams

See how business applications, data, and third-party services support and are impacted by the use of AI.