From building a compliance system on Power Automate to exploring AI for RFPs, smaller firms are leveraging new tools once reserved for larger players
By Corinium Global Intelligence
Chris Taylor honed his leadership style at big firms like State Street and Northern Trust. Now, as COO of the boutique firm CG Asset Management, he is using the breadth of experience he gained at these major players to supercharge a smaller team.
Taylor, who is a Founding Advisor to AssetOps, spoke to Corinium about extracting value from cloud-based, low-code tools, optimizing due diligence workflows, and aligning with vendors for maximum impact.
How has the COO’s role changed in the decade you’ve spent at CG Asset Management?
It’s become more technical, and over time it’s moved away from high-level governance and more into detail. At the same time, it’s become more strategic – deciding how to get the best bang for your buck when investing in technology, and deciding whether to allocate resources to people or to tech.
Automation has always been about working smarter and getting more done with fewer errors. In the last decade, we’ve seen incredible new opportunities open up for smaller firms, which is where I’ve been focused for the last five years. These simple, powerful tools have given us access to the same advantages that used to be reserved for big companies.
Which opportunities in particular are now available to boutique firms like yours because of the cloud?
Running your own data center is expensive, not just the hardware, but the maintenance and support that come with it. The cloud changes that. With Microsoft’s scale, you get access to virtually unlimited computing power and storage at a very competitive price point. You can even create a hybrid setup that combines the best of on-premises and cloud resources.
With things like Power Automate, Power BI, Power Apps – suddenly it's all low code. You can show it to people and ask, “what would you like to achieve?” We can build something internally as a wireframe and hand it over to developers to make it secure and scalable. In the past, you'd have been speaking to consultants and IT professionals, spending money before even getting to a proof of concept.
Today’s low-code tools encourage experimentation and creativity without requiring a huge upfront investment of time or money. One of my earliest projects with it was about five years ago; a compliance system for quarterly attestations, PA dealing, and the recording of gifts and inducements. With little more than a few tutorials and Power Automate, we built an entire workflow system with compliance questions, approval flows, reminders, and contract management in an incredibly short space of time. Once that had been done, it was handed off to the developers to be scaled and hardened for production. That kind of rapid prototyping just wasn’t possible before without a large development team.
The cost saving for that system alone is tens of thousands of pounds a year compared to an off-the-shelf application. There’s no recurring cost – just time upfront and some consultancy to build it. Once you’ve started, you can build more incrementally. We’ve built mobile apps which allow receipts or invoices to be recorded at the point of purchase and automatically logged in our compliance and accounting software. You simply take a picture and fill in some standard fields and the software does the rest; it saves huge amounts of time at month end.
Which other areas do you think are going to have a big impact on buy-side operations in the next few years?
One area is building AI directly into processes. We’re not doing that at the moment but we are speaking to service providers that do, so we can benefit from their investment. Right now, AI is in a growth phase with a lot of players, and I think there will be a period of consolidation. I’ll wait until the market stabilizes and then look at which tools integrate best with our processes.
Some of the most prevalent use cases at the moment, especially for boutiques, are Requests for Proposals (RFPs), Due Diligence Questionnaires (DDQs), and research. These are well-suited to AI, provided the source data is regularly reviewed, effectively structured, and appropriately secured. I appreciate that there are a lot of exciting developments in this space, but there are also a lot of people using a lot of resources and achieving very little. AI is not a silver bullet; there needs to be governance to ensure answers are correct, useful, and in context.
I like to think of AI as an employee. You give it a job description, review its work and then provide feedback. Consider where the AI needs training, where it falls short, and whether it is completing its job description to the standards you expect.
Can you talk about some AI use cases you’re working on at the moment?
I’ve experimented with combining multiple language models and having them build on each other’s outputs. With appropriate guardrails established, we can layer AI tools which use policies, presentations, meeting notes, and previous responses to generate responses that can be accessed internally. We have an AI usage policy which determines exactly what the output can be used for, and this policy will evolve as we get more comfort in the results.
Where AI still struggles is context. If you ask, for example, about the biggest areas of operational risk, it might include valid responses but also list risks which are not relevant in the context of the question. This highlights the need to give AI a helping hand by keeping the data well structured, using metadata tags, and refining the prompts.
AI is very complex and building your own would require a large budget. There are a number of AI LLMs which are developing at an exponential rate already, so our time is better spent ensuring that our prompts are effective at generating what we need.
How do you assess whether a potential AI use case will add value?
I look at what other people are doing, success stories where it’s been done well. Then I see how this aligns with our strategy and infrastructure. The value comes from the efficiencies it brings and this can, in the short term, be measured quite easily. I ask: What would it cost to have a task done manually and what would be the cost of developing or licensing a solution?
I am still slightly skeptical of the claims I hear, but I am always open to being proven wrong. Many claim they have the tools to fix a wide range of problems, but there seems to be a disparity between these claims and the end users experience. It is often said that implementation can be a journey, and both sides have to develop or enhance their systems to tailor the output.
If you have a large team or you need something bespoke then that is fine, but if you’re paying a large amount of money for a solution or service then you need to partner with a firm that knows where it's going, how to get there and has the skills to take you along with them.
I will always try to adapt our internal processes to fit the standard operating model of a key service provider. I take comfort in knowing that when developments are available to other clients, it’s been road tested and rolled out for us as well.
What motivated you to become a Founding Advisor for AssetOps, and what do you think differentiates this initiative?
I think the fact that it’s a buy-side event being put together by the buy-side means we can shape the agenda around the things that are important to us right now. We’re having these conversations all the time, we know the challenges we have, and it’s helpful to bring those to an open forum where everyone can share their experiences.
Some of us have gone further than others, some have gone down a route and realized it was wrong, and some of us are marching steadily towards that same mistake with a sense of confidence that implies we are doing the right thing. In coming together in an open forum, we can share our experiences, learn from each other, and become more effective in our roles.
One of the reasons I wanted to join early and help to shape it was to make sure there’s representation from firms of all sizes. Boutiques in the City with ten to 50 employees are often underrepresented at mainstream events, and yet the collective Assets Under Management is incredibly large. The challenge of a boutique can be just as relevant to a large asset manager; it just needs to be seen through a different lens.
Chris Taylor will be speaking at AssetOps Digital London on November 13. View the full agenda and register here.