Webinar recording now available.
Watch the full session below.
The recording above captures the full discussion, including the exchanges and nuances that a summary cannot fully convey. We recommend watching the session in full if you are working on any of the following challenges:
We have summarised the key themes from each section of the discussion:
LPAs, side letters, and the terms that govern everything: LPAs, side letters, and the terms that govern everythingThe limited partnership agreement is the foundational document of every fund. But in practice, LPAs are rarely standard. Side letters create bespoke carve-outs. Fee structures diverge. Hurdle rates, preferred returns, carried interest splits, distribution waterfalls, all of these can vary from one investor to the next within the same fund.
Karen Sands opened by describing what “more precision” actually means in practice: LPAs cover fee mechanisms, exclusion rights, ESG provisions, transparency obligations, and manager-level ODD requirements. Side letters — once genuinely exceptional — have become a standard item in every fund close. Steven Merriett confirmed this from the LP side: OPERS negotiates bespoke arrangements across a portfolio of 250 funds, with increasing demands from their oversight board to protect beneficiaries. The result is a web of contractual obligations that back-office teams must track, reconcile, and execute accurately; increasingly, without the headcount to do it manually.
The operational burden this creates is significant. Managing multiple LPAs with different fee structures, hurdle rates, and distribution terms requires precise, document-level logic not generic templates. When that logic lives in spreadsheets or is held only in the institutional memory of a finance team, the risk of error and inconsistency compounds with every new fund and every new investor.
What a waterfall calculation actually looks like when it goes wrong: One of the more pointed moments was Oliver’s account of what waterfall calculations look like at a firm still running them on Excel. The typical reality: a model built eight or more years ago by someone who has since left the firm. Data is pulled from multiple fund administrators in different formats. The original logic has never been independently verified. The model gets extended over time — Oliver called it the “sauce” problem: the same recipe with new ingredients added, but no one checks if the original sauce was ever right. Auditors sign off. The model runs. And it can produce a confident, internally consistent, materially wrong answer. He also addressed a common misconception directly: some firms are now trying to run waterfall calculations through AI tools. This does not solve the problem. Language models are probabilistic — they cannot guarantee the 100% accuracy and full reproducibility that legally binding distribution calculations require.
Errors in waterfall calculations can result in incorrect distributions, GP clawback exposure, LP disputes, and findings in operational due diligence. The more complex the economics, the higher the probability of a confident, internally consistent, materially wrong result.
qashqade's role in this context is to replace that exposure with a system that takes the LPA logic, enforces the correct sequencing, and produces a fully auditable calculation trail for every distribution event.
When hundreds of investors replace a handful of institutions: Managing three or four institutional LPs with bespoke LPAs is operationally demanding. What happens when that number becomes hundreds as it does in retail feeder structures, wealth management platforms, and democratised access vehicles?
Karen made the structural complexity concrete: when people talk about a fund, they typically mean a single entity, but the reality is often six to eight legal entities, each with its own terms, reporting requirements, and jurisdictional obligations. Most GPs operate across multiple jurisdictions with a global investor and investment base. Bespoke terms work at low volume; without proper infrastructure, customisation without standardisation becomes unsustainable as volume grows. The panel also raised the talent dimension: Steve noted that skill requirements are changing to suit both technology and AI, and that OPERS has had to build new job ladders to retain people with the right hybrid capabilities.
Continuation vehicles, evergreen funds, and the problem of inherited complexity: Continuation vehicles and evergreen funds add another dimension to the challenge. Unlike traditional closed-end funds, these structures have no fixed end date and no clean waterfall moment. Distributions are ongoing. Capital flows in and out continuously. And critically these vehicles often arrive already carrying fee and term conflicts from prior vehicles they are designed to bridge.
Jordan noted that GP discretion clauses in LPAs make continuation vehicles particularly tricky to administer — and that getting this right requires building strong client relationships from day one, bringing GPs into the room early to agree on terms rather than discovering conflicts after the fact. Oliver added that qashqade had to deliberately develop software capable of handling evergreen structures, because the challenge is qualitatively different from traditional closed-end funds. The lesson from eight years of experience: there is no shortcut to handling this complexity. You learn by doing, and by accumulating experience across many different structures. The audit trail must be persistent, fee calculations must be continuous, and the logic must account for obligations inherited from prior vehicles.
Automation: beyond speed: A recurring theme in this section was the question of what genuine automation actually delivers — beyond simply doing the same thing faster. The panel's view was clear: the value is not primarily speed. It is accuracy, consistency, and defensibility at scale.
An automated allocation system that guarantees LPA-driven logic does not just calculate faster. It calculates the same way every time, across every investor, with a documented methodology that can be reviewed, replicated, and defended. That is a qualitatively different outcome from a faster spreadsheet.
Operational weakness as a reason not to re-up: One of the most significant exchanges in this section involved the LP perspective on operational due diligence. The question put to the panel: at what point does operational weakness become a reason not to re-up, regardless of investment performance?
The answer, from the LP on the panel, was unambiguous. Operational infrastructure is a due diligence signal. A GP that cannot demonstrate a clean, auditable trail for fee calculations and distributions raises questions about governance, risk management, and institutional readiness. For large allocators these are not secondary concerns.
A formal ODD finding about inadequate distribution controls travels. Within LP investment teams, to co-investors, to the consultants who advise multiple allocators. The reputational dimension is not separate from the operational one.
The final section of the webinar addressed the question that is on every operator's mind: addressed the question that is on every operator's mind: where does AI actually fit in the management of fees, incentives, validation, and audit-readiness?
The panel's view was measured and grounded. AI is adding real value in specific, well-defined applications: document review and LPA extraction, anomaly detection in large datasets, drafting and summarisation in reporting workflows, and surfacing patterns across fund data that would take analysts significant time to identify manually.
Oliver’s opinion? Where AI is not yet fit for purpose is in the core calculation layer. Waterfall calculations are legally binding. They require deterministic, reproducible results. Language models are probabilistic by architecture. They cannot provide the institutional-grade audit trail, the version control, or the guaranteed reproducibility that regulated fund managers require.
The panel was clear on this distinction: AI as a co-pilot in fund operations is genuinely useful. The more interesting near-term opportunity, the panel suggested, is using AI to make purpose-built systems more accessible and more intelligent, surfacing insights from structured data, accelerating LPA ingestion, and improving the quality of reporting outputs, while keeping the calculation logic in a controlled, auditable environment.
From operational efficiency to competitive differentiation: The earlier conversation had focused on automation as a way to reduce risk and manage complexity. This section asked a different question: is there a bigger opportunity here?
Various visions of the future emerged. Steve described a world of genuine LP-GP data coordination: LPs and GPs collectively wasting enormous time on phone calls and emails that could be replaced by a portal-like experience — data available when needed, in the format required. He sees this as the next major evolution in the investor relationship. Karen identified taxonomy standardisation as the critical missing link: too many different names for the same concepts across the industry means that even the best technology cannot achieve true end-to-end solutions without either a common language or robust mapping between systems. Jordan shared his view on identifying the partners that are taking the right next steps with AI to be a long-term partner. Oliver offered a prediction: no single software will ever cover every need in this ecosystem. Instead, in four to five years, the best specialist tools will have been embedded into larger platforms — with an AI layer on top — and qashqade is positioning to be part of that ecosystem. The firms that treat operational infrastructure as a fundraising asset, rather than an administrative cost, will have a structural advantage as this consolidation plays out.
The investor portal, the reporting layer, the ability to provide real-time visibility into fees, distributions, and fund performance, these are LP-facing products. And the quality of those products increasingly influences LP decisions about where to commit capital.
The firms that recognise this shift early and that treat their operational infrastructure as a strategic asset, not just a back-office function, are the ones that will have a structural advantage in fundraising over the next five to ten years.
Oliver Freigang, CEO, qashqade: Oliver leads qashqade, the allocation management platform purpose-built for private markets. qashqade enforces LPA-driven logic, automates waterfall calculations, and delivers auditable, scalable infrastructure for fund managers, fund administrators, and LPs.
Steven Merriett, Assistant Director, Investment Accounting, Operations & Compliance, Ohio Public Employees Retirement System (OPERS): Steve oversees investment accounting, operations, and compliance at one of the largest public pension funds in the United States, with deep experience in LP-side operational due diligence and fund governance.
Jordan Rothberg, Head of Product, North America, IQ-EQ: Jordan leads product for North America at IQ-EQ, one of the largest global fund administration groups, with direct experience managing the operational complexity of diverse LP bases, feeder structures, and alternative fund formats.
Karen Sands, COO, Federated Hermes Private Equity: Karen oversees operations at Federated Hermes Private Equity, bringing a COO-level view on the practical challenges of managing LPA complexity, investor reporting, and operational scale across a diversified private markets programme.
