Muamer Cisija is cofounder and president of Symphony, a global software design and development company.
AI is not the future—it is already shaping our present.
According to a recent Forbes survey, AI and machine learning represent the top tech spending area in 2025, capturing 42% of IT budgets. Moreover, an NTT survey conducted by WSJ Intelligence shows that 89% of global CEOs consider AI crucial for future profitability.
My recent conversations with clients, particularly private equity (PE) executives, have also shown that leaders have already decided that they should implement AI. Whether it’s through AI-powered code analysis, predictive analytics in retail or automating documentation in healthcare, companies understand they can find value in targeted applications of AI.
What they want to know now is how to do so effectively. Having worked with a variety of companies navigating this shift, I’ve observed key patterns for the broader industry. Drawing from real-world use cases I’ve seen across sectors, here are five critical lessons for applying AI to transform legacy software and unlock growth.
1. Start small and solve specific pain points.
The best way to introduce AI is not through broad, disruptive transformation but rather targeted, incremental changes.
Recently, one of my company’s partner mid-market PE funds faced talent erosion after acquiring a healthcare company. The immediate challenge was documentation gaps and reliance on a few individuals who understood legacy codebases. Using automation tools tailored to legacy systems, the team quickly generated documentation, and this phased approach allowed them to embrace AI organically, building trust and preparing them for more advanced AI adoption.
According to Boston Consulting Group, 74% of companies haven’t seen “tangible value” from using AI, but they also found that companies that lead in AI implementation invest strategically in a few high-value use cases before scaling. In other words, starting with small wins can help your organization see value quicker and understand how to implement AI elsewhere.
2. Purpose-built AI delivers superior results.
In my experience, most PE and enterprise companies start their AI journey by adopting generic, off-the-shelf developer tools. While beneficial, these solutions are limited because they’re not tailored to specific contexts.
In contrast, AI solutions that focus on specific issues, like converting stored procedures, tend to deliver greater efficiency. By focusing their AI efforts on specific legacy issues, I’ve seen companies achieve automation levels of around 60%, drastically cutting the typical refactoring time and costs by 50% to 80%.
As a recent Harvard Business Review article explains, “Context is everything” when it comes to AI. Organizations are most likely to see value from their AI journeys when the tools align with the organization’s specific needs.
3. AI should amplify, not replace human expertise.
One significant lesson emerged from client feedback: Technology alone isn’t enough—client collaboration matters just as much. While deep-dive documentation and code analysis are a valuable starting point, successful modernization requires early engagement to uncover root problems, align on priorities and build a clear roadmap tailored to the client’s business context.
As analysis progresses, AI and human expertise take on distinct but complementary roles across three phases. In due diligence, AI conducts deep analysis guided by human-set parameters. In planning, AI proposes optimization paths—cost, time or market—while humans shape the final roadmap. Execution then follows, with AI handling up to 80% of tasks when automation is feasible. Throughout, humans remain in control, ensuring direction and alignment.
When AI is combined with expert oversight, the resulting insights often go beyond automation, driving greater efficiency, employee satisfaction and a shift from reactive maintenance to proactive innovation.
4. Scalability and robust support are non-negotiable.
According to the Forbes research mentioned above, scalability and post-purchase support rank as the most critical criteria for selecting AI and machine learning solutions. Yet scaling AI is not just a matter of adding infrastructure—it’s about designing with adaptability from the outset.
Based on my experience, scalable AI implementations often rely on cloud-native architectures, modular APIs and well-documented pipelines that can evolve as business needs change. This also requires embedding AI within core operational processes rather than treating it as a side initiative. This allows AI to grow with the company—not become a brittle bolt-on.
On the flip side, common pitfalls include over-customization that limits portability and failure to plan for cross-functional buy-in and governance. Flexibility isn’t just technical—it’s also organizational. Companies must invest in training, build cross-team workflows and plan for continuous feedback loops to keep AI initiatives aligned with real-world conditions.
5. Proactivity beats reactivity.
Perhaps the most profound shift AI-driven modernization enables is transitioning companies from reactive to proactive. As AI rapidly transforms industries, organizations must anticipate shifts rather than respond after disruptions occur.
For instance, AI-enabled predictive analytics allows companies to detect and respond to market trends in real time. Instead of scrambling after an event happens, retail companies can proactively adjust inventories and marketing strategies.
But, as the MIT Sloan Management Review notes, 92% of companies they surveyed felt that “cultural and change management challenges” have been the top barrier to becoming data- or AI-driven. AI can give your organization the tools to do so proactively, but companies must ensure their teams are ready to act on AI’s insights.
The AI-Powered Edge For Business Growth
AI modernization is more than a technological upgrade—it’s an essential competitive strategy. But achieving its full potential requires more than adopting tools or automating tasks. Leaders must guide these efforts with a long-term, business-aligned vision. Start by asking:
• Are we targeting the right high-impact problems?
• Do we have the processes and talent to scale AI?
• Are we using AI to enhance—not replace—human expertise?
With only 26% of AI projects seeing “tangible value,” according to the Boston Consulting Group survey cited above, achieving success requires clear goals, scalable infrastructure and committed executive leadership. Just as critical is a cultural shift, one that encourages cross-functional collaboration and reimagines legacy workflows.
Whether you’re a PE investor or a CTO, these lessons offer a starting point to shape strategy and guide internal conversations. The companies that move boldly and intentionally today will lead tomorrow.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
