Investment Ops Automation Will Give You Back Date Night — But You've Still Got Work in the Morning

The tolerance for error in investment operations is nil. A bad NAV, a missed corporate action, a settlement break that doesn't get caught in time isn't a rounding error. It's a client's trust, a buy-in and punitive fees, a firm’s reputation. Everyone who has sat inside an operations team for any length of time has felt that particular dread: The numbers don’t tie, it's 8 P.M. and you’ve got places to be, and it needs to be fixed before anyone goes home.

I spent the first 14 years of my career in that seat. And the clients I supported were rarely the simple ones. The most sophisticated managers — the ones pushing into new asset classes, new structures, new strategies to find their edge — are, by definition, the ones creating the most operational complexity. That's not a coincidence. Complexity is where the returns are. It's also where the risk lives.

That complexity brings an inevitable hunt for expertise. Someone must become the person who understands how this new instrument settles, how their economics reconcile, how to account for their margin. You become the subject matter expert because you were voluntold, or you were asked for help and you didn’t step back quickly enough. You're the one everyone leans on precisely because you can do the manual work nobody else can.

That tension — expertise as both the solution and the trap — followed me when I moved out of operations and into relationship management, and then into sales. I assumed the themes would change. They didn't. Whether I was sitting across from a director of ops explaining a break, or a CFO explaining why we were the right partner for their next fund launch, or a COO painting the art of the possible with a data transformation project, the same conversation kept resurfacing in different clothes: Technology only matters if you have the right people using it.

For as long as growth has bred complexity, the answer has been the same, and easy to defend in a budget meeting: Add more people. Like Warden Norton said in The Shawshank Redemption, "More bars, and more guards."

More funds, more asset classes, more jurisdictions, more investor reporting requirements: headcount, headcount, headcount. Stretch the same manual processes across a bigger surface area, and when the stretching wasn't enough, hire again. It worked, in the narrow sense that the numbers eventually tied out and the reports eventually went out the door. But "it worked" is a low bar, and it hid a cost that never showed up on the org chart.

The hidden cost was this: Heroics were mistaken for good process. When a team pulls an all-nighter to get a close done, get an investor letter out, or catch a break before it becomes a client-facing problem, everyone breathes a sigh of relief and moves on. Nobody stops to ask why it took an all-nighter. Surviving the close isn't the same thing as the close being well-designed. But when survival is the daily measure of success, no one has the time or incentive to ask the harder question.

That was the model for a long time because, frankly, there wasn't a better option. The tools available couldn't handle the volume or the complexity without a human checking every step. Which brings us to what's changed. The math that used to work — throw more people at more complexity — has run out of road.

Consider what "complexity" means today versus 15 years ago. Arcesium now supports more than 400 asset classes across traditional, alternative, private markets, and digital assets. Crypto and digital assets trade around the clock, with no natural pause for a close. Investor reporting expectations have compressed from quarterly narratives to daily dealing and the need for near real-time transparency. Regulatory reporting requirements multiply every year, in every jurisdiction a fund touches.

The people capable of doing this work well don't want to do it this way anymore. The best operations talent — the people who understand the nuance of a new instrument or a complex fee structure — increasingly won't tolerate a career built on all-nighters and manual checking. They leave for roles that use their judgment instead of their patience. So, firms relying on the old model face a double bind: The work is getting harder, and the people willing to grind through it manually are getting scarcer.

The good news is that cloud-native, AI-driven platforms can now genuinely streamline reconciliation across trades, positions, cash, and account balances; manage treasury and collateral workflows with real oversight; automate complex fee structures and investor allocations; and streamline regulatory reporting with automated form completion and data validation. This is the first time the tooling has matched the complexity of the business.

Here's the catch, though, and it's the part I think the industry gets wrong most often: Automating a workflow is not the same as removing the need for a control. Done right, automation doesn't eliminate the control, it relocates it. The control becomes an exception-based review process, an automated tolerance check, an audit trail, lineage tracking that shows exactly how a number was derived. The person's role shifts from, "Catch the error by staying late" to "Design the system so the error surfaces itself and know what to do when it does."

The hours that used to go into producing a number should go into interpreting it, flagging what a NAV move actually means for a client's exposure, spotting a pattern across funds that a system won't connect on its own, having the conversation with an investor before they have to ask the question. Operations, done well in this new model, isn't a cost center you keep alive with long nights. It's a source of alpha the rest of the firm doesn't have anywhere else, because the operations team sees the whole trade lifecycle.

A platform that can automate reconciliation and reporting is only as good as the team that knows what the outputs mean, knows when to override an exception, and knows how to explain a number to a client who's asking a hard question at a bad time. Technology without that judgment doesn't reduce risk — it just moves the blind spot from, "We didn't check," to "We didn't understand what the system told us."

If I could go back in time and talk to the version of myself scratching his head over a Where’s Waldo game of “find the break,” I wouldn't tell him to work less hard. I'd tell him the hours were going into the wrong thing. I’d probably also tell him to buy a lot of bitcoin.

Luckily for our collective sanity, grueling hours aren’t the only option anymore. The firms that get this right over the next decade won't be the ones that automated the most, or the ones that cut headcount the fastest. They'll be the ones that took the hours technology freed up and reinvested them into judgment, oversight, and client relationships. They will never stop asking whether the control is still there.