As artificial intelligence reshapes software development, one long-standing question is being asked again with renewed urgency: should organisations build their own systems, or buy existing solutions?
In this discussion featuring Theodora Lau of Unconventional Ventures and Matthew Barnard of BBD Software, the answer is not as straightforward as technology headlines might suggest.
AI has undeniably lowered the barrier to entry for building software. Development is faster, prototyping is easier, and the cost equation is shifting. What was once considered too expensive or time-consuming to build is now being reconsidered. Projects that previously defaulted to SaaS solutions are increasingly back on the table as viable custom builds.
But speed alone does not solve everything.
As both speakers emphasise, enterprise environments — particularly in financial services — require far more than rapid development. Accuracy, security, scalability, and regulatory compliance remain critical. In industries dealing with money, being “almost right” is not enough. Systems must be robust, auditable, and reliable under pressure.
This is where the distinction between building and buying becomes more nuanced.
Off-the-shelf solutions offer maturity, embedded expertise, and proven functionality. They often represent years of intellectual property and refinement. However, they can also come with compromises — organisations may only use a fraction of the functionality while paying for the full product.
On the other hand, custom builds provide flexibility and the ability to tailor solutions precisely to business needs. With AI accelerating development, that option is becoming more attractive. But it introduces its own challenges: long-term maintenance, governance, scalability, and the complexity of bringing software into production.
The conversation also highlights an important shift in where differentiation occurs.
Core systems — particularly in areas like accounting or payments — tend to remain standardised. Differentiation is increasingly found in customer experience, interface design, and how services are delivered, rather than in the underlying infrastructure itself. AI enables this by allowing organisations to personalise and adapt experiences more quickly.
Looking ahead, one of the most significant trends is the move towards agent-based systems and more structured enterprise AI adoption. Rather than relying on ad hoc “vibe coding,” organisations are beginning to formalise how AI is integrated into development processes, focusing on accuracy, governance, and repeatability.
At the same time, the industry is likely to see changes in pricing models, greater regulatory scrutiny, and ongoing debate around risk, control, and transparency.
The conclusion is not that one approach will replace the other. Instead, the future lies in balance.
Building and buying will continue to coexist — with AI shifting the boundary between them, but not removing the need for careful, context-driven decision-making.
