The Drawdown talks to Henry Lin, founder and CEO of Linnovate Partners, about how AI technologies will transform fund administration.
What is your strategic vision for how AI will transform the role of the fund administrator and the services you provide to private market funds?
At Linnovate Partners, automation is deeply embedded in our DNA because that’s where we focused from the start. So strategically, AI is a natural evolution of what we’ve been building, a way to further enhance efficiency and insights for our clients. That said, we recognise that the adoption of AI can present challenges, especially for more traditional administrators who may need to adapt their existing processes.
Across the industry, advanced technology is becoming more accessible but implementation isn’t always straightforward. We’ve been fortunate to integrate AI in ways that help transform traditional back-office functions into more strategic, value-added services. For example, we use AI to intelligently interpret complex documents, automate labour-intensive tasks and generate predictive analytics – all in real time. This allows us to provide deeper, faster insights to our clients while maintaining accuracy.
We’re also leveraging AI to improve communication with investors, moving beyond templated responses to more dynamic, tailored interactions. Another key focus has been portfolio data capture, particularly for fund-of-funds. Historically, this required significant manual effort, but AI now enables us to systematise these processes while maintaining precision.
Ultimately, we see AI as a tool to elevate the entire industry, helping everyone to deliver greater transparency, efficiency and strategic value to private market funds. We’re excited about the possibilities and look forward to seeing how these advancements benefit all stakeholders.
So does that mean a cost saving for clients?
It’s not necessarily about saving costs upfront. I often use this analogy: If I take a flight from Singapore to London, the journey typically takes 11 hours. If I could get you there in five hours, you’d expect to pay a premium, right? The same applies to AI – it may not always translate to immediate cost reductions but it delivers a far better service.
For clients, it’s about long-term value. Manual processes used to carry risks like human errors and delays, but AI improves accuracy, speed and consistency. Over time, these efficiencies lead to a stronger, more competitive pricing model. And this is just the beginning; it’s an ongoing process and we’re continuously evolving to deliver even greater value.
When is an AI truly necessary and in which scenarios can advanced automation sufficiently address operational needs?
For us, automation is the foundation – it efficiently handles all repetitive operational tasks like NAV calculations, reconciliations and standard reporting. This baseline of automation is critical because it creates the stability and efficiency needed to explore more advanced solutions like AI.
When we are looking at the different AI tools to deploy, we are trying to identify areas where the automation falls short, when tasks become very demanding. For example, with complex document interpretation, with very long documents or side letters. We have an LLM integration that helps with reading and extracting information from those documents.
Another area is cap-table documentation, where previously people had to read each individual paragraph. Now we use AI to speed up the process of extracting the relevant information, so people now spend time reviewing rather than extracting.
For fund-of-funds or venture capital firms, they might be investing into 80 portfolio companies. There is no way, traditionally, that you are managing that documentation without manual processes. But now with complex documentation interpretation tools, you can start to process things a lot quicker.
And as I mentioned before, when it comes to LP requests, there is no way you can use standard templates to interpret those questions. But that is where AI can come in. With the right training, it will recognise the question that the LP is asking and help pick the right template to respond to the request.
The key is balance. Automation handles the fundamentals brilliantly, while AI steps in to solve higher-complexity challenges. Together, they empower teams to work smarter, not harder, but neither is a one-size-fits-all solution.
Where does AI provide unique value beyond what traditional automation can achieve in fund administration?
I would say the unique value of AI is the way it can make these automations smarter, so they don’t have to wait for prompts. When a request comes in, it automatically knows what to do next and informs you what its reply would be. If you agree, you click send.
It is something that sits on top of the automation, which you have already built. If you don’t have the automation, you don’t have the foundation to effectively use AI or to know which models to use.
We see it as building an administration agent to sit on that automation foundation in order to get things done quicker. That is how we see the technology being used and what we are working on.
What is your roadmap for the next two years on future AI-driven services and functionalities?
Our approach to AI is grounded in practicality, security and incremental improvement rather than chasing trends. At Linnovate, we carefully assess each process to determine where AI can add real value while maintaining robust data security and compliance. We don’t adopt AI for the sake of innovation – instead, we focus on core areas where it can drive meaningful efficiency without compromising accuracy or confidentiality.
For some of the work we do, AI implementation just isn’t necessary. Drafting an annual financial statement, for example, is impossible to do with current AI technologies because of the level of interpretation required.
We are just trying to be very practical with what we can do with AI. We look around the market and a lot of peers are starting to explore all these different AI initiatives. But for us, we are just focused on a few core areas.
For example, portfolio data capturing. We could be handling 300 to 500 quarterly datasets during the peak season. Before, it would take us days and nights for two weeks to process all that, but now we have cut that process time by more than 50%. This is just the beginning and we’re trying to streamline further.
Another area is using advanced document intelligence for extracting information, in areas such as KYC and AML. Looking at how we can extract the passport information or other information. These are quite labour-intensive processes, so developing these tools can be really helpful. A standard optical character recognition technology can read the documents but it can’t understand changes to them, so introducing AI has improved data extraction and saved our teams a lot of time.
We are also working using genAI for advanced reporting. Previously, reporting was based on a set of templates but now we are starting to allow our users within our internal team to interact with a chatbot and use that chatbot to generate those reports. Templates are fixed whereas a chatbot plugged into AI is able to respond to LP requests much quicker.
Those are the main focuses for the next two years and, as a business, we always upgrade our technology every two years. We have done that consistently, adding new tools both for clients and internally.
The way I think about AI now is like having a supervisor. If you are new to a project and need help understanding what you need to do, you can ask it and it will give you suggestions and tell you where you can find out more information. It is a way of expanding what you are doing without massively expanding your resources.
Fund administration remains a people business and you can not convert it entirely to not relying on people to solve problems. But our philosophy is to be practical and always consider what more we can do with technology.
This article was originally published in the The Drawdown Fund Admin Report 2025.
