The Global Investment Leaders Club gathering on Investing in Autonomous Legacy Transformations brought investors, founders and family office representatives into a practical discussion about where AI and deep tech are already changing legacy systems. The conversation moved away from broad excitement around AI and focused instead on workflow adoption, defensibility, infrastructure, talent and measurable return on investment.
Modernization is no longer only about adding AI tools on top of existing processes. It is about redesigning how work is organized, how data moves, and how old systems can become faster, more autonomous and more useful. A USA-based investor said adoption is visible but still early: “We are beginning to see implementation of artificial intelligence in workflows in our business and also in our portfolio companies, but it is very early days.” He pointed to private equity portfolio companies as one area where AI could create value, especially where highly paid manual work can be reduced or improved.
Praharsh Shahi, VC at Mount Judi Ventures, gave a more operational view of this shift. He described how even smaller investment funds are already using AI to rebuild internal workflows, noting that “within a week” an engineering intern was able to recreate workflow tools after reviewing vendor demos. For him, this showed that AI is “not only changing the way big enterprises are thinking about their entire workflows,” but also disrupting smaller organizations. He also connected this to India’s rise from back-office support into global capability centers, backed by a growing deep tech talent pool.
The discussion also showed that investors are becoming more careful about what counts as a real AI opportunity. For John Sharp, Managing Partner of Hatcher+, the investment potential lies less with AGI, and more with specific models built for enterprise deployments. He said, “Most consumers view AI as a sophisticated search engine. Creating value for investors from AI is going to require the development of specialized, enterprise-class agentic services - and an understanding of the complexities involved in deploying AI inside sophisticated, proprietary systems. This is especially important for financial activities, where accuracy is absolutely critical."
Pasi Pohjala, Founder and CEO of ATG Consulting, added that investors are now asking a sharper question when looking at AI agents: “Is this a solution looking for a problem?” He argued that startups become more attractive when they can clearly explain the cost savings, efficiency gains and operational speed they create. This point helped ground the discussion in commercial discipline rather than AI excitement alone.
Anneliese Sound, Managing Director at Future Potential Management, focused on the organizational challenge behind AI adoption. She argued that legacy transformation cannot happen if companies keep their old decision structures. “The classical way people are organized in companies will not work anymore,” she said, adding that businesses need faster decisions and more authority delegated down the line. For her, the bottleneck is not only technology, but leadership knowledge. “We are still learning, yes, but we have to learn faster,” she noted.
Other investors pointed to deep tech opportunities beyond software alone. Alessandro Zago, Senior Venture Investor at VU Venture Partners, highlighted physical AI, robotics, industrial automation, autonomous vehicles, data center infrastructure, rare earth materials and space technology as major areas of interest. He described physical AI as a field covering “robotics, industrial automation, humanoid robots,” while also pointing to greener data centers and new infrastructure for transportation and space.
Art Norins, Partner at TLG Investment Partners and TLG Venture Partners, brought the conversation back to enterprise adoption. He said his team focuses on AI for Fortune 500 companies as well as small and medium-sized businesses, where applied AI can improve work processes in a way customers can measure. “We see huge opportunities with small medium businesses and large Fortune 500 companies to deliver them real value using AI,” he said.
Vishal Arora, Founder and Managing Partner at PanCosmic Capital, broke the agentic AI opportunity into four layers: application, platform, infrastructure and energy. He noted that data management will become a sharper concern in the AI lifecycle, while energy consumption inside data centers will require software led optimization. His comments reflected a wider investor view that AI’s next opportunities sit not only in applications, but also in the systems that support them.
The summit also broadened the meaning of legacy transformation beyond enterprise software. James McDowall, Founding Partner at Arcanum Capital, argued that AI will need a financial transaction layer of its own, asking, “What money is AI going to use?” He pointed to stablecoins and blockchain infrastructure as part of a deeper modernization of legacy financial systems. Elizabeth Addonizio, Investor, also highlighted blockchain’s role in supply chains, critical minerals and healthcare, where digital audit trails can improve transparency between providers, payers and other market participants.
The summit also included a closer look at companies trying to bring autonomy into legacy environments. A founder presented an AI native IT platform designed to resolve enterprise IT issues before they disrupt operations. He framed the shift as moving from “static data and analytics based systems” to platforms with “dynamic intelligence and execution.” He also described the future as being driven by “autonomy as opposed to automation, outcomes as opposed to dashboards, and invisible AI as opposed to noisy analytics.”
Still, participants were clear that AI does not remove the need for discipline. Ian Valentine, Private Investor, challenged the idea that “software is dead,” saying it is “not 100% true” because AI-generated systems can produce poor results without experienced people guiding the process. He warned that companies still need “disciplined product development and specification processes” if they want AI tools to work reliably.
For investors, this was one of the clearest examples of AI moving from promise to measurable enterprise value. A founder said his project aims to move IT issue resolution from less than 5% automatically resolved today to more than 80% autonomously resolved, creating lower operating costs, higher productivity and an ROI at least 10x higher than competing platforms.
The broader message of the summit was that autonomous transformation is not a simple software upgrade. It requires better data control, stronger operational design, real platform depth and clear economic impact. The most attractive opportunities are not AI companies in name only, but businesses using AI and deep tech to modernize legacy sectors where cost, speed, reliability and resilience truly matter.
