Taking on the Obstacle of Technical Debt to Take Advantage of AI
AI’s emergence has created opportunities to enhance customer experiences and streamline operations through automation. However, technical debt, which takes the form of outdated systems, software and processes, can get in the way of the successful adoption of AI, hindering digital transformation efforts.
New research compiled by Insight, in partnership with Foundry, an IDG company, confirms as much. It shows how prolific AI is. For example, 92% of IT decision-makers who responded to the survey, The Path to Digital Transformation: Where IT Leaders Stand in 2024, said they are currently leveraging or testing AI/machine learning technology.
Technical debt meanwhile ranked as one of the top challenges inhibiting innovation (36%; Gaps in technology skills and knowledge was first at 44%). Another factor that makes digital transformation follow-through inconsistent is back-and-forth decisions between traditional data centers and public clouds like Amazon Web Services (AWS), Google Cloud Platform (GCP) and Microsoft Azure.
There was justifiable initial excitement over public clouds, but some of the difficulties include:
- Inadequate assessment of workloads to determine their suitability for cloud migration
- Inadequate cost analysis of digitization initiatives and operational expenses
- Inadequate training and development of IT staff to adapt to the cloud
- Delays in re-platforming or refactoring to leverage cloud-native functionality and platform/software-as-a-service solutions (treating clouds as traditional data centers)
Businesses frustrated by the resulting delays may be tempted to bypass IT altogether instead of addressing the root cause. However, that can lead to more technical debt in the form of shadow IT. Much like interest on financial debt, technical debt compounds. Innovation slows as a result. One specific federal agency spent $800K in one year just to maintain infrastructure a decade old. It reached out to Insight to modernize its IT in favor of a more strategic, cost-effective — and sustainable — approach.
To fully leverage data’s potential and facilitate the adoption of AI, organizations must modernize their aging infrastructure. However, it calls for careful planning. Where do they begin?
1) Take inventory.
The first step involves conducting a thorough assessment of the organization’s current IT estate. This establishes a baseline of what it must work with to create a plan of action.
Begin by focusing on compute resources. For VMware users, this may involve exporting inventory information with the RVTools utility that connects to vSphere. Other platforms have similar tools available.
Next, move on to storage and network resources. Storage discovery is typically performed using vendor-specific tools like the NetApp Inventory Collect Tool (ICT) and Pure1 from Pure Storage. Network inventory tools include Nagios XI, SolarWinds, Datadog and others. More comprehensive asset management packages combine these capabilities with IT Service Management (ITSM) methodologies.
2) Evaluate the IT estate.
Organizations also must determine who owns each asset. This is hard work, especially in larger companies. Identifying application and server owners can lead to uncertainty, particularly with custom apps. They may not be properly documented, or the apps could be owned by people no longer with the organization (who failed to properly transition ownership to someone else upon their departure).
Regardless, systems must be determined as actively in use and by whom. Addressing these issues can reduce technical debt right off the bat. Other factors include:
- Apps running on outdated infrastructure
- Multiple deployed versions of the same software (and/or multiple competing solutions)
- End-of-service penalties, expiring contracts, and ending warranties or support services
One of the most crucial evaluations is each system’s Total Cost of Ownership (TCO). As systems age, TCO increases and penalties and contract expirations can tip the balance against the status quo.
3) Address technical debt.
Assessments up to this point should uncover opportunities to invest in new technologies while reducing reliance on outdated systems and software. The next step is to create a remediation plan.
If you’re like many of Insight’s clients, you may come across systems and apps ready to retire, freeing up resources. Now you can consolidate onto supported platforms without having to refactor apps.
The more technical debt retired this way, the easier it is to focus on replacing outdated hardware, upgrading software, and updating obsolete processes and procedures. It helps to identify change champions and gamify initiatives and milestones. Reward teams for identifying and decommissioning rogue systems and eliminating shadow IT accounts and software.
4) Stay out of technical debt.
With a modernized IT infrastructure, companies are better equipped to drive innovation, deploy new services and solutions like AI, and create net-new value. Still, the ever-present challenge is to avoid incurring technical debt. Companies would end up right back where they started.
Adhering to established frameworks helps. Examples include Information Technology Infrastructure Library (ITIL) or ITSM, Control Objectives for Information and Related Technologies (COBIT), The Open Group Architecture Framework (TOGAF) and DevOps.
Also, it’s time to (re)consider how IT resources are consumed and whether the team is viewed as a partner to the business or merely overhead. If it’s the former, organizations can empower IT to create value through innovation, differentiation relative to competitors, improvement to employee experiences and more.
Other strategies to stay out of debt:
1. Update inventory information regularly, at least on a quarterly basis. However, it’s even more beneficial to implement a real-time inventory and IT governance system.
2. Ensure Change Management (CM) processes are appropriate for supported platforms. Legacy CM designed for physical infrastructure with three to five-year lifecycles is insufficient for managing virtual infrastructure like public clouds. In this environment, systems can be deployed and decommissioned quickly via Infrastructure as Code.
3. Use tagging and automation with policy to accurately identify owners and regularly apply updates.
4. Implement cost-management systems or tools that evaluate TCO. Tie it back to budgeting processes to keep spending in line with expectations.
5. Regularly re-evaluate platforms. As new capabilities and pricing models emerge, you’ll probably want to adjust levels of investment in hardware/software and data center/public cloud solutions.
6. Align IT investments more closely with business goals. This could involve revising the traditional three to five-year IT renewal strategies often dictated by Capital Expenditure (CapEx) depreciation schedules. Companies can make better use of public cloud investments as a result.
Technical debt is a challenge that organizations can manage effectively with the right preparation, tools and investments. By addressing technical debt, organizations reap numerous benefits and avoid the headaches associated with aging systems. The most significant reward though is the ability to leverage new technologies like AI to drive innovation and gain a competitive edge in your industry.
For more information on how to reduce your technical debt and get the most out of AI, visit insight.com or call 1.800.INSIGHT.
