By Ashish Shrestha, Senior Solutions Consultant at MAIA Technology: Artificial intelligence is moving quickly from experimentation towards implementation across the hedge fund industry. For managers, the potential applications are broad. AI could support investment research, portfolio analysis, operational process automation and the interrogation of increasingly complex datasets. But as firms move from exploring individual use cases towards embedding AI into everyday workflows, a more fundamental question emerges:

Is the operating environment beneath AI capable of supporting what the fund is asking it to do?

For hedge funds, this matters particularly because complexity can accumulate quickly. A manager may launch with a relatively straightforward operating model, only to add strategies, instruments, funds, counterparties, data sources and service providers as the business develops. Technology stacks often evolve alongside this growth. New systems are introduced to solve specific requirements, integrations are built between them, and manual processes fill the gaps. AI does not automatically remove that complexity. In some cases, it exposes it.

AI depends on the infrastructure beneath it

An AI-enabled workflow is only as effective as its ability to access reliable information and interact with the systems where investment and operational activity actually take place. Consider a hedge fund operating across multiple asset classes, prime brokers and counterparties. Portfolio, order, execution, risk, cash and P&L data may exist across different platforms and arrive at different points in the investment lifecycle. Before an AI application can meaningfully interrogate that information, the data needs to be accessible, consistent and current.

If information must first be extracted from multiple systems, transformed or reconciled, AI inherits those dependencies. If underlying platforms operate through batch processes, the picture being analysed may not represent the portfolio now. And if connecting a new application requires proprietary interfaces or extensive development, every new AI use case risks becoming another integration project. None of these are new operational challenges. AI simply raises the importance of addressing them.

From AI experimentation to live hedge fund workflows

There is also an important distinction between proving that an AI application works and trusting it within a live investment environment. A proof of concept can operate within a controlled dataset. Production is different. If AI is expected to interrogate positions, exposures, cash, P&L or trading activity, the quality of its output becomes inherently connected to the quality, lineage and timeliness of the underlying data.

That places greater importance on architectural principles such as real-time data, interoperability, API-first connectivity and cloud-native infrastructure. For hedge funds, these foundations also have implications beyond AI. They influence how quickly a manager can introduce a new strategy, accommodate greater trading complexity or scale the business without continually expanding its operational footprint.

But AI readiness is not simply about making more data available to a model. It is also about controlling what AI can access and defining what it should be responsible for. Through permission-aware APIs, investment firms can create a governed route between frontier AI models and real-time investment data, ensuring that access reflects the permissions of the authenticated user.

Hedge funds can already leverage AI to interrogate exposures, explain P&L movements, summarise cash and margin pressures, triage operational exceptions or draft investor reporting commentary. In each case, the business benefit is not that AI replaces the underlying investment infrastructure. It is that users can access trusted information faster, understand it more clearly and act with greater confidence.

Growth can expose technology debt

This is where the AI conversation becomes part of a much broader operating-model discussion.

A fragmented technology environment can function adequately at one level of complexity. But as trading volumes rise, new instruments are introduced or the organisation expands, the cost of maintaining interfaces, reconciliation processes and manual workarounds can increase. The result is technology debt: resources are increasingly directed towards maintaining the existing environment rather than improving it.

Modernising the underlying architecture can therefore serve two purposes. It creates a stronger foundation for emerging AI-enabled workflows while also addressing existing constraints around data accessibility, interoperability and operational scale.

MAIA was designed around these principles. Its cloud-native, API-first architecture brings portfolio management, trading, risk, compliance, IBOR and middle-office workflows into a unified environment, reducing fragmentation and creating greater consistency across the investment lifecycle.

That architecture also creates a controlled foundation for AI interaction today. Users can leverage frontier models such as Claude and Codex to interact with MAIA, with its permission-aware API framework governing access to investment data.

Crucially, the underlying architecture remains the source of truth and calculation layer. MAIA’s IBOR continues to aggregate investment data and perform deterministic calculations across positions, cash, exposures and P&L. AI can sit above that foundation as a query, explanation and report-drafting layer. Rather than asking an AI model to replicate critical business calculations, it can be used to interact with trusted information produced by the investment infrastructure beneath it.

AI readiness is operational readiness

The next phase of AI adoption within hedge funds is unlikely to be determined solely by access to increasingly sophisticated models. Those capabilities are developing rapidly and becoming more widely available. The greater differentiator may be the environment into which they are deployed.

Over time, AI could become an operational command layer for hedge funds. It could help teams identify the most important exceptions, explain the likely cause of breaks, anticipate cash or margin pressure, monitor close processes, draft reporting commentary and surface operational risk before it becomes visible through traditional reporting. The result is a more scalable control environment, where growth in strategy complexity does not require a proportional increase in manual oversight.

For hedge fund COOs and CTOs, that means AI readiness should not sit separately from conversations about the investment operating model. Data architecture, system interoperability, real-time processing and technology debt all influence what AI can ultimately do in practice. Equally important is the distinction between the investment infrastructure responsible for trusted data and deterministic calculations, and the AI layer used to interrogate and explain that information.  The question is therefore becoming less about whether AI will transform hedge fund operations. It is whether the operating model is ready to let it.