.svg)
.svg)
1. Hiring won't fix complexity. Fund structures, LP numbers and regulation are growing faster than ops teams. The firms pulling ahead invested in technology, and size has little to do with success.
2. LPs want data they can use. That means clean, exportable data delivered consistently, not 50-page PDFs. Even basics like net IRR still aren't reliably available.
3. AI is only as good as your data. Being able to use unstructured data is a real step forward, but it only pays off with clean foundations and a human checking the output.
4. Building in-house doesn't scale. It works while things are simple, but the case for buying gets stronger as complexity grows.
Last week, our team was on the ground at SuperReturn CFO/COO in Amsterdam, one of the most important gatherings in private markets. Across panels, roundtables and corridor chats, one theme kept coming up: private markets firms are taking on more complexity than their operations can handle. Below, we summarize our learnings and suggestions looking ahead into 2027.
The problem everyone is talking about
Private capital markets continue to grow exponentially. More funds, more structures, more LPs, more regulatory requirements across multiple jurisdictions simultanesouly. But the operational infracture supporting that growth hasn't kept up, in many cases.
The default response has been to add headcount. More people to manage more complexity. But as one panellist put it plainly: that approach is expensive, error-prone, and fundamentally unscalable.
The firms pulling ahead aren’t the largest ones. They’re the ones willing to invest in technology and engage proactively with operational innovation. Size turned out not to be the differentiating factor - mindset is.
LPs want more, and they want it usable
If there was one message that came through loudest across every session, it was this: LP demands have grown significantly and show no signs of slowing.
When a capital call or distribution goes out, IR teams are now fielding hundreds of questions simultaneously. LPs are more sophisticated than ever, using GP data within their own portfolio management systems for comparison - which means they need it clean, flexible, and exportable rather than buried in a 50-page PDF.
The ask is both simple and demanding: give us more underlying data, in a format we can actually use, consistently, and on time. Basic metrics like net IRR are still not reliably accessible across the industry. The firms winning on LP relationships are the ones treating investor reporting not as a compliance exercise but as a relationship asset.
The AI moment is real - but so are the caveats
AI was everywhere at SuperReturn - as a topic, a solution, and a source of genuine excitement and healthy scepticism in equal measure.
The biggest unlock cited this year is the ability to have unstructured data sitting alongside structured data. Capital call context, deal descriptions, portco updates (information previously buried in PDFs and email inboxes) is now accessible and actionable in ways it simply wasn’t twelve months ago. For an industry that still traffics heavily in PDFs, this is genuinely transformative.
Firms that invested in clean data infrastructure early are already seeing the compounding benefits. Those who didn’t are finding that AI tools are only as good as the data underneath them, which is why data hygiene kept coming up as the foundational conversation beneath every AI discussion.
The honest caveat that resonated most: for the foreseeable future, you still need a human in the loop to sense-check AI that is confidently wrong. The firms getting this right are using automation to eliminate grunt work and free their teams for the judgement-intensive work that actually requires human expertise.
Build vs buy
Across multiple sessions, the headcount versus tooling debate came into sharp focus. When operational requirements are simple, firms try to build in-house. But as complexity grows, that approach breaks down, and the firms that recognised this early are now significantly ahead.
One example that stuck: 120 hours of coding time to build an LP dashboard and quarterly reporting automation that now runs continuously, freeing the team entirely from repetitive manual construction and allowing them to focus on refinement and insight.
The question being asked in 2025 is no longer whether to invest in technology. It’s whether you have the data hygiene and operational foundations to make that investment actually pay off.
What this means for the road ahead
Evergreen fund structures are proliferating - over 200 launched last year alone - bringing new operational complexity around liquidity management, real-time fund administrator coordination, and valuation consistency. Regulatory requirements across the UK, Europe, Asia, and the US are adding simultaneous strain. And LP expectations of white glove service are extending even to forecasting and scenario planning.
The firms navigating all of this well share a common characteristic: they stopped trying to solve operational complexity with people alone, and started building technology infrastructure that scales with their ambitions rather than against them.
Our take
What we heard in Amsterdam matches what we’re building at Atominvest: one operational layer for private markets firms that removes fragmentation, automates repetitive work, and gives GPs and LPs the data they need in a format they can use. Get in touch if you’d like to see how it works.
.png)
