Where technology meets judgment: Aztec insights in Private Funds CFO report

Scaling with technology doesn’t have to mean losing the human touch, say Aztec Group’s Akbar Sheriff and Jason Adams in their recent interview with Private Funds CFO.

For private markets fund managers, the challenge has often been how to make back-office processes more efficient. Rapidly evolving technology means efficiency and high-touch service no longer conflict. Originally published in Private Fund CFO’s Automation, Systems and Technology Report, Akbar Sheriff, chief client success officer and chief operating officer at Aztec Group, and Jason Adams, our chief technology officer, explain how managers can streamline routine workflows while preserving human judgment, access and relationship quality.

How are firms deploying technology to improve operations?

Akbar Sheriff: Nowadays LP expectations are high, as they navigate multiple relationships with different managers and fund services providers. They expect greater access to information, intuitive processes and basic tasks to be completed without repetition. They want accurate, secure, timely and seamless interactions with technology, for instance during onboarding, NAV reporting and when addressing cross-jurisdictional regulatory requirements. They don’t want tech for tech’s sake or simply more data.

However, when judgment matters, or an LP has a complex request, or is seeking additional information, or requires assistance, they still expect direct access to the GP. Using a high touch, high-tech model, technology elevates the client experience for the LP, while team members at the manager can proactively nurture those relationships and become a trusted partner.

Jason Adams: There’s huge variability and little standardization. Firms continue to grapple with highly variable data, use cases, nuanced fund structures and constantly evolving asset classes. Concurrently, the volume and time sensitive nature of LP requests are increasing. As the industry scales, fund managers need technology to automate the routine, less-cognitive workflows to keep pace with LP requests and maintain a competitive advantage, while offering specific human expertise and capabilities. This approach allows firms to be more effective and valuable to their clients.

What does a manager need to have to operate a high-touch, high-tech model?

AS: In the past, firms have focused on solving specific problems with separate point solutions. Particularly with AI, there is a risk of ending up with a collection of apps or isolated use cases that are difficult to scale across clients and lack supporting infrastructure. While encouraging the use of AI, firms need to monitor use cases to prevent duplication of effort, consider using existing AI platforms and measure effectiveness.

They also need a coherent plan for AI adoption and a governance framework. All this requires them to think strategically about business priorities and client impact. For example, reducing time spent on a single element of investor onboarding is not as effective as reviewing the entire process to identify greater efficiencies.

JA: For a manager, it makes sense to implement AI to tackle large volumes of unstructured documents, including PDFs and e-mails, and to automate rules-based workflows. The value isn’t in any individual tool or platform; it’s in the connection between them. Data fragmentation and quality are the biggest risks in scaling a firm’s data architecture. Historic reliance on point solutions is compounded by the segmentation of asset classes and results in different teams using different data, which then requires reconciling before it can be accessed across the business.

Establishing a digital ecosystem is crucial for GPs to manage increasing complexity as they launch new funds or expand into new asset classes or jurisdictions. First, they need to standardize their data, structure it into a schema or a warehouse, then integrate workflow layers to access that information efficiently and establish a data governance model that ensures data quality.

Once foundations are in place, managers can focus on delivering complex capabilities such as AI-augmented workflows, or modern user and LP experiences. That is likely to define operating models of the future: blending intelligent automation, connected data and specialist human expertise. Over the next five years, the firms that scale most effectively will be those that can industrialize routine processes while preserving service quality, judgment and trust.

Are firms buying or building infrastructure to support their ecosystems?

JA: Most CFOs are trialing AI pilots and building proofs of concept, but few are building production grade solutions. The foundational elements of agentic AI are expensive and difficult to execute and maintain. It makes sense to invest in industry-specific solutions that build on existing AI infrastructure developed by hyperscalers such as Microsoft, AWS or Google, including data pipelines and integration frameworks that connect different point solutions, accompanied by an AI policy and monitoring of output quality.

AS: Firms need to think about how they maintain their edge. Technology alone is no longer a competitive advantage; the differentiator is how firms combine technology with experience, judgment and service. They also need to retain control of their data and avoid dependence on any single provider. Firms need a technology roadmap and should empower teams to innovate and develop use cases that are responsive to market demands, built on a well governed, scalable foundation.

How are firms measuring AI impact?

AS: One key measure is whether AI is contributing to broader business success, and how clearly the firm can connect AI to those outcomes. On the client side, we track inputs like AI training, and outputs such as user numbers, which applications are being used and how frequently. The more meaningful question is whether AI helps the firm make better use of its context layer.

JA: Clients are using AI in three ways: using large language models and generative AI to write code and build products improving capabilities; using AI agents to improve operations and processes; and augmenting their offering to clients, by embedding agentic capabilities into a unified client experience.

One basic way to measure AI adoption is by how much a firm is spending on tokens. A more sophisticated way is how AI is used in workflows, tracking efficiency gains, quality improvements and novel insight generation. This method is more aligned to traditional product value measurement, helping the firm understand where AI is making a truly positive impact.

What do LPs expect in terms of cybersecurity and operational resilience?

JA: While there’s been LP pressure on firms to adopt AI, they remain very sensitive to cybersecurity and broader operational resilience risks. The cyber threat is evolving with more frequent and complex cyberattacks, including impersonation and fully autonomous attacks.

But investors are also focused on governance, lifecycle management, controls and the resilience of the wider operating model. These concerns are reflected through in-depth LP due diligence questions around cybersecurity and technology governance.

Investors have moved beyond a tick box approach and are now asking about infrastructure, AI policy enforcement, monitoring, proactive threat modelling, red team exercises that mimic malicious actors, and the controls in place to manage technology safely over time. No one can ignore the severity of the threat.

If you want to find out more, please contact us below.