It is time for asset managers to embrace modularity

Every asset manager wants the same four things from their technology stack: fewer contracts. Fewer handoffs. Fewer portals to log into. A single place to look when something breaks.

The shorthand for all of it is consolidation, and the industry is acting on it. The 2026 InvestOps report, a survey of 200 senior operations executives at firms above $10 billion, found 58% naming vendor and platform consolidation as a top initiative. Modernizing architecture and data infrastructure came second, at 54%.¹

The instinct is right. An integrated platform removes duplicate processes, cuts handoffs, and gives a team one trusted view across the operations lifecycle. But there is a cost. When you commit to one vendor’s front-to-back platform, you commit to that vendor’s product decisions. If they build what you need, great. If they take the product somewhere you did not want to go, your options narrow. And if something better appears for one part of the operation, you have a new problem.

Say an exception management product arrives that can investigate and route breaks your current system handles manually. You should be able to adopt it. If your operating model was designed around one tightly integrated platform, adopting it means reconfiguring the surrounding workflows, rebuilding integrations, and reopening the architecture and the contract. The specifications that made the platform attractive are the same ones that now make it expensive to change.

That is the ordinary case. The bad case is deciding the platform itself is wrong, which buys you a multi-year replacement, a major investment, and a long disruption to the business. So the status quo offers a choice between passing on a capability that puts you at a competitive disadvantage, or taking it and starting a project to fit an external component into an architecture not built for one. Neither is attractive.

BCG’s 2026 asset management report puts industry margins broadly flat around 30% since 2010, with costs growing slightly faster than revenue, 5.4% against 5.1%, and management fees down roughly 23% over the same period.2 Some of the most interesting opportunities to optimize are in the middle and back office.

And the way to do it is to discard the false dichotomy above.

Firms are not wrong to want fewer disconnected systems. McKinsey’s July 2025 work on the economics of asset management found that 60 to 80% of technology budgets go to running the business rather than changing it, that firms often fail to fully decommission legacy systems after modernizing, and that managers work in siloed data environments with no fit-for-purpose front-to-back platform available to them.³

So then, the problem isn’t too many vendors so much as too little coherence.

A modular platform is a set of components that can change independently while still behaving like parts of one operating model. A single-vendor platform is coherent because everything inside it was built to fit. A modular one is coherent because an overlay carries context and data from one component to the next, whoever built the component underneath.

The overlay is where the operating model lives. Beneath it, components may come from different vendors with different data models and different logic. Above it, an operations team sees one workspace: risk and compliance, trade validation, accounting, settlements, exceptions.

The efficiency of one platform. The capability of the best tool for each job to be done.

As much as research analysts like to prophecy, no one knows which technology will be best in five years because the future growth of knowledge is unpredictable. The vendor with the strongest reconciliation product today may be unseated by a company — or a technology — that does not exist yet.

A firm should not have to make a five-year bet every time it buys a piece of software. And yet that is what happens when the operating model is built around a single platform.

Modularity creates optionality. A firm can trial something, keep what works, replace what does not, and do it on its own schedule rather than a contract’s.

The schedule matters more than it used to. Citisoft’s 2025 transformation research puts typical change programs at around 17 months for firms under $100 billion, with enterprise platform implementations commonly running 12 to 24 months.4 Capability is arriving at a faster pace. Between February and April 2026, Anthropic, OpenAI, and a little startup called Google released seven models between them, roughly one every two weeks, and the wider software market has compressed the same way.5

Firms will need to make more changes to their operating models, not fewer. They will need to adapt individual workflows continuously, and to respond to technologies whose significance was not obvious when the underlying systems were chosen.

That is clearest with AI, which is moving from something people use to something that runs across the investment lifecycle. BCG estimates 25 to 35% cost reduction available over three to five years, and notes in the same breath that most firms are still piloting rather than transforming.6 The distance between those two facts is an infrastructure problem.

Today, a person is the connective tissue between systems. They move between tools, gather what is relevant, work out what needs to happen, and then execute it. An agent can take on much of that. But only if it can absorb context and act across systems that belong to different vendors.

Take a reconciliation break. Resolving it means comparing a portfolio position against the accounting record and custodian data, reviewing the underlying transaction, checking the rules and tolerances that apply, deciding whether the break is legitimate, and then routing or resolving it. An agent doing that work has to know what has happened, what is outstanding, which system is the authoritative book-of-record, what rules apply, what action it is permitted to take, and what happens downstream if it acts.

It needs data, yes, but also context, workflow, and control.

The complexity and cost of assembling that information increases when every system in the operation only functions inside its own ecosystem.

Fragmented systems have always created friction for people. They are about to create more friction for machines, and it will cost more. A person who has to open three applications to close a break is slow. An agent that cannot move between them cannot do the job at all.
 

A modular operating model only works if the technology underneath can behave as one. Interoperability is the precondition. The overlay design is what makes it useful: one place where exceptions, tasks, approvals, SLAs, documents and controls are coordinated across the systems and teams involved.

The components underneath will change. Oversight cannot. Every action, workflow and user stays traceable, permissioned, and subject to one set of governance rules.

And in the age of AI there is a second requirement: the controls that make automated action safe. A system might identify an exception, recommend a resolution, and start the workflow. In regulated operations, recommendation, approval and execution cannot blur together. Higher-risk actions still need human review, sign-off, and clean auditability.

Guardrails by default. Traceability, audit trails, permissioning, defined human-in-the-loop control. Product outcomes.
 

This is our thinking behind Optima, Linedata’s investment operations orchestration and oversight layer. It connects existing systems through APIs, unifies workflows, centralizes exception management, monitors SLAs, runs configurable rules, and enables controlled automated action inside a governed environment. Monitoring and alerting are table stakes. Optima recommends the next step, triggers approved actions, and resolves eligible exceptions.

Linedata reported production results at launch: up to 80% less manual effort across document-driven workflows, a 50 to 60% reduction in financial reporting cycle time, real-time SLA visibility, and more than 70% of data exceptions resolved automatically.7

In practice that looks like NAV oversight with rules-based checks, SLA tracking and digital sign-off. A reconciliation exception centralized, assigned and routed rather than chased across email and spreadsheets. Trade settlement exceptions identified, routed and resolved with a full audit trail. Document processing that extracts information into downstream workflows while keeping human review where it belongs. Because it is an API-based overlay, those workflows can run across Linedata and third-party systems. Firms are not asked to replace the accounting, NAV or reconciliation platforms they already rely on.

The difference is not what the product does, it is what it already knows. Optima is built for middle and back office investment operations rather than enterprise workflow in general. A NAV window, a settlement fail, a provider SLA, a maker-checker gate: these are not configurations bolted on; they are the product’s design pattern.

This is where investment operations is going. One operating model, built on components that can change at the pace of the market. Where the layer you standardize on and the components you buy underneath it are two different decisions, and only one of them should be hard to reverse.

For more information on Optima and how to set up a modular architecture for your investment operations, chat with our team.
 

Thomas Dadmun leads Linedata’s global product management team, developing portfolio management and investment operations systems for asset managers. Previously, he founded Point Focal, where the company explored natural language processing (NLP) and alternative data for trade signaling and risk management. He has worked in data strategy and product analytics at State Street and began his career as an equities trader. Thomas also taught in Northeastern University’s Master’s in Analytics program and enjoys working at the intersection of markets, data, and technology.