Last February, nearly $285 billion was wiped from software stocks in a single trading session.
Most people interpreted the moment as another wave of AI exuberance colliding with stretched valuations. But beneath the volatility was something more structural. Investors suddenly realized the traditional SaaS model may no longer be as defensible in a world where AI agents can navigate workflows without relying on the application interface itself.
That distinction matters more than most industries currently appreciate.
I recently read a piece from ThoughtLinks CEO Sumeet Chabria outlining what may be one of the clearest frameworks yet for understanding how agentic AI changes enterprise software architecture. The argument is simple but profound: the interface is no longer the moat.
For the last twenty years, enterprise software won by controlling workflow. If you wanted to manage CRM activity, financial planning, portfolio analytics, client communication, compliance, or reporting, you logged into separate systems and manually navigated each environment. The software experience itself was the product.
Historically, firms faced a binary technology decision: buy software off the shelf or spend enormous capital building custom infrastructure internally. For most advisory firms, building was unrealistic. So the industry defaulted toward accumulating SaaS subscriptions and stitching workflows together manually through people and process.
Agentic AI changes that equation.
Firms can now maintain existing systems while building proprietary intelligence and orchestration layers above them, creating a hybrid model that combines the scalability of purchased software with the strategic advantage of owned infrastructure. That shift is significant because agentic AI changes where value resides inside the technology stack.
When AI agents can move across systems, retrieve context, synthesize information, execute workflows, and coordinate actions autonomously, the value shifts away from the interface and toward the intelligence layer orchestrating the activity. In other words, firms no longer need to operate entirely inside vendor-defined environments. Increasingly, they can build intelligence architectures that sit above existing systems and unify fragmented workflows into a cohesive operating model.
This is where wealth management becomes particularly interesting. Despite the trillions of dollars managed across the industry, most advisory firms still operate on fragmented infrastructure assembled one subscription at a time. CRM systems disconnected from portfolio systems. Planning tools disconnected from communication systems. Research disconnected from client engagement. Institutional knowledge trapped inside PDFs, inboxes, meeting notes, and siloed applications.
The result is operational fragmentation disguised as modernization. Most advisors do not suffer from a lack of software. They suffer from an excess of disconnected software requiring constant human coordination.
That fragmentation becomes exponentially more problematic as firms move upmarket into increasingly complex client relationships. Entrepreneurs, executives, family offices, and multi-generational households do not live inside clean workflow categories. Their financial lives span concentrated equity, private investments, operating businesses, liquidity events, estate structures, philanthropy, tax strategy, and intergenerational planning simultaneously. The advisory model increasingly requires contextual intelligence across all of it.
Traditional SaaS architecture was never designed for that level of orchestration. It was designed for workflow specialization. That distinction is becoming critical.
The firms that benefit most from agentic AI will not simply automate repetitive tasks faster. They will rethink how intelligence flows across the enterprise altogether. Instead of asking employees to manually navigate systems, they will create environments where AI continuously synthesizes context across the organization and surfaces the next most relevant action automatically.
That is not incremental productivity improvement. That is a fundamentally different operating model. More importantly, it changes the strategic question firms should be asking right now.
The question is no longer simply: Which software should we buy? It is increasingly: Which capabilities should remain vendor-driven, and which forms of intelligence should firms own themselves?
Because agentic AI lowers the barrier between buying and building. Firms no longer need to replace every system to create differentiated infrastructure. They can layer proprietary intelligence, orchestration, workflows, and contextual reasoning on top of existing applications without rebuilding the entire stack from scratch.
That hybrid model has enormous implications for wealth management because the firms with the strongest client relationships already possess the most valuable asset in the AI era: proprietary context.
Not public data. Not generalized models.
Context: Client history. Behavioral patterns. Relationship dynamics. Planning complexity. Decision-making preferences. Timing sensitivities. Family structures. Business ownership. Tax exposure. Liquidity events. Personal objectives.
That is the real strategic asset. The firms that successfully operationalize that intelligence will not simply become more efficient versions of today’s advisory practices. They will become structurally different businesses.
In the next article, I’ll explore what that shift demands - and why it’s not another SaaS tool.
Humans Lead. Agents Scale.