The investment management software industry is in the middle of an Artificial Intelligence (AI) arms race. SaaS vendors are racing to add AI capabilities to their applications, with each month bringing new chat interfaces, recommendation engines, and workflow assistants. Yet most are destined to underdeliver.
The AI era will not simply separate vendors with “better AI” from vendors with “worse AI.” It will separate application providers from solution providers.
What’s the difference?
While application providers view the world through the narrow prism of their own software, solution providers use AI to solve business problems that span multiple applications, workflows, and data sources.
Application providers own an application.
They embed AI into that application. They create a few AI-powered workflows and pat themselves on the back, convinced that their clients' workdays revolve around those workflows, much as medieval monks believed the sun and stars revolved around the earth. But a CRM activity is not a client meeting, and there is no "make investment decision" button inside a portfolio management system.
The outcomes investment firms care about are inherently cross-functional—not operating within a single application or dataset. From Microsoft Teams to their OEMS, Google Search to Salesforce, Bloomberg to Excel, users spend their days stitching together information from disparate systems and navigating complex, cross-functional workflows to deliver the outcomes they desire.
The Problem with Locking AI Inside a Walled Garden
Imagine a portfolio management vendor creates an AI-powered assistant for preparing for investor meetings.
Ahead of the meeting, the assistant provides a summary of portfolio performance, holdings, and investment activity. Is this helpful? Yes. It saves the user time by gathering information from various reports and dashboards. Is it complete? Not even close.
To name just a few examples, the investment manager also needs:
- Prospectus information
- Investment research perspectives
- Recent communications with the client
- Service history and any outstanding issues
- Overall client sentiment
- Key stakeholders and relationships
- External events affecting the client
Any one of these factors can fundamentally change the context of the meeting and the relevance of the insights drawn from the data. Operating within a silo, the portfolio management vendor's AI solution inevitably reaches a ceiling.
This siloed approach may be a little smarter, and the user may save a step or two, but in the context of the broader workflow, very little has changed. Most of the work still sits with the investment manager, crisscrossing applications to assemble the context needed to derive meaningful insights.
The New Control Point in Enterprise Software
For decades, enterprise software has been organized around applications. Users logged into a CRM to manage client relationships, a portfolio management system to monitor investments, an accounting platform to reconcile positions, and countless other systems to perform other specialized tasks. The application was both the interface users engaged with and the control point through which information, workflows, and decisions were managed. AI changes this dynamic.
As enterprise AI becomes more capable, users increasingly interact with AI first and the underlying applications second. Instead of navigating multiple systems to gather information, users ask AI a question. Instead of manually orchestrating a workflow, they describe an outcome. Enterprise AI determines how that outcome can best be achieved.
In this model, AI is more than an assistant. It is the connective tissue that helps users access information, execute workflows, and take actions across multiple applications. The underlying applications continue to perform critical functions, but increasingly as systems of record and systems of execution rather than primary points of engagement—capabilities behind the scenes rather than destinations in their own right.
The strategic implication is profound. As enterprise AI becomes the primary interface between users and technology, the most important question for software vendors will no longer be how intelligent their application is, but how effectively it participates in a broader intelligence ecosystem.
The Fundamental Mistake Many Vendors Are Making
Most vendors sell software applications, so they assume the future will consist of many AI assistants associated with each of those applications—a CRM assistant over here, a portfolio management agent over there. But no CIO, COO, portfolio manager, or operations leader is asking for ten or twenty separate AI tools.
What they want is holistic AI experience that can:
- Understand their entire operating environment and learn from organizational knowledge
- Reason, execute, and connect structured and unstructured information across applications
- Surface insights proactively and deliver real business value
In other words, they don’t want another application, but intelligence that spans the enterprise, connecting systems, knowledge, workflows, and decisions.
Until recently, this would have been impossible. However, major advances in model performance, the emergence of Model Context Protocol (MCP), and the rise of multi-agent architectures and reusable agent skills are bringing the contours of Enterprise AI into focus.
This creates an entirely new paradigm for software vendors. Historically, SaaS companies were built around managing and monetizing single application experiences. Enterprise AI shifts the value proposition from application-centric to ecosystem-centric.
The Temptation to Own the Ecosystem
Recognizing the importance of ecosystem-wide intelligence, some vendors will attempt to become the ecosystem themselves. While this sounds attractive in theory, in practice it is extraordinarily difficult.
Building enterprise intelligence requires more than adding AI to an application. It requires two foundational capabilities: a data foundation that unifies and governs information across the enterprise, and an AI platform that can reason across that information, orchestrate actions, and operate safely at scale.
Enterprise Data Platform Capabilities
- Data architecture, governance, and productization
- Integration, APIs, streaming, and event orchestration
- Data automation, observability, and platform operations
- Identity, security, authorization, and auditability
- Cloud and ecosystem integrations
Enterprise AI Platform Capabilities
- Agent orchestration, governance, and lifecycle management
- Models, MCPs, memory, and context management
- Human oversight, traceability, and compliance
- Integration with enterprise workflows, tasks, and events
This is not simply a feature roadmap. It is an entirely different business.
Few software vendors possess both the technological breadth and organizational mandate to deliver these capabilities. For most, the more sustainable opportunity lies not in owning the enterprise ecosystem, but in becoming a trusted participant within it.
What This Means for Investment Managers
For the last two decades, enterprise software has been defined by applications. Because software vendors primarily delivered value through their applications, success was often measured by application adoption and utilization. Solutions, however, are optimized for outcomes. One maximizes seat licenses; the other maximizes business value. For many years, those objectives were broadly aligned. In the age of AI, they are increasingly diverging.
For investment managers, this shift carries an important implication: technology decisions made today will determine whether they benefit from the transition to Enterprise AI or spend years funding architectures that increasingly work against it.
The question investment managers should ask prospective technology partners is no longer, "What AI features do you have?" The question is, "How does your technology participate in my enterprise intelligence ecosystem?"
The firms that thrive in the AI era will not be those with the most applications, the most workflows, or even the most agents. They will be those that enable intelligence across the enterprise—turning information into decisions, decisions into actions, and actions into outcomes. A divide will form between vendors that can provide that and the rest.
