What it needs to do, and what to ask the vendor before you commit
TLDR: A portfolio research system defines where your research is stored, who can access it, and whether you can trace the thinking behind a position you took months ago. These key aspects separate a good platform from an expensive mistake: everything that impacts an investment decision should be in one place, the right people should see the right data, a vendor who already works with firms like yours, and the ability to ask your own data a question in natural language.
Most investment teams inside hedge funds and asset managers do not run their entire process in one system alone. They rely on data vendors, risk providers, models in spreadsheets, research in email, and notes from the sell side.
This works well most of the time but when a PM or an investor has a question that spans across multiple systems at once, it becomes a project. Answering a question about a position’s history and the decision process that led to its sizing sometimes takes a day of digging, and the answer depends on how well all the information surrounding the investment decision was preserved. This is why it’s critical for a modern fund manager to have all their investment information accessible on one centralized, secure system.
Why a cloud-based portfolio system
A cloud-based system keeps the work in one place instead of scattered across file folders, third-party tools, and email. It is a centralized repository of all the information you need to run your investment management process.
When all your risk and performance data is stored in one central system, the complex questions take a few minutes to answer instead of hours of pulling data from different sources. Being in the cloud simply means that an analyst with the right entitlements can access it while traveling or working from home, the same way the PM can work in it from the office. The whole team sees the same numbers at the same time, without emailing versions of a file to each other. Updates arrive continuously, so the platform is current when you need it. This is now table stakes for a modern portfolio management system.
What it needs to do: the features
First of all, it should help you securely access all your investment data in one place, on demand from anywhere. Fundamentals, estimates, prices, risk factors, and increasingly, alternative data. An alt data set provides higher ROI if the signal is easily accessible to the analyst or PM doing work on a position. They often need to act quickly and do not have extra time to find the login to an alt data platform or pull the model where that data is stored. We recently wrote about this topic here.
A great system lets you work with risk, attribution, factor exposure, notes and memos relevant to your positions, all in the same environment. Understanding risk is critical, but it must be measured in the context of the future return expectation and the conviction that the analyst and PM have in the position.
It needs to be secure and only give the right people access to the right data. Entitlements should be robust enough to differentiate by role and provide access to the employees who need that data. A platform that can only grant all-or-nothing access tends to limit the information that is shared there, which defeats the point of having it.
Finally, it must be future-proof. Your platform should outlive technology cycles and function as the backbone of your investment management process for the long term. The practical test today is whether it exposes your data to AI tools through an open standard rather than a closed integration. Our clients are now able to use their favorite AI models like Claude to securely access their portfolio data. More on that here: https://equitydatascience.com/articles/eds-mcp-layer-ai-investment-system-of-record/
- "Understanding risk is critical, but it must be measured in the context of the future return expectation and the conviction that the analyst and PM have in the position. "
What to ask the vendor: the partnership
The product demo is the easy part for the vendor as they have had time to prepare and polish their pitch. These questions can help you go deeper and tell you a lot more about how good a fit the vendor is for you.
Do they work with firms like yours?
Ask which of their clients look like you in size, strategy and structure. A vendor serving twenty funds that function like you has worked through many versions of your challenges, and the product reflects that understanding and experience. It’s also important to ask which strategies the product is best for. Systematic investment strategies have a very different workflow from fundamentally driven ones and the tools will be different.
Ask who on their team has sat in your seat and how they approached the same challenges you face today. Subject matter expertise on the vendor side is the difference between a feature set you will use and something that looks good at first but has no long-term value inside your org.
Do they innovate?
Look at what shipped in the last twelve months and also ask about their roadmap. Are they thinking about how technology is changing the way investors run their business and their portfolios? A system that helps you manage your investments is core to your organization and should be future-proof; it is not something you want to change often. It is hard enough to keep pace with technological advancement in the age of AI, but if the vendor is not committed to innovation, your tools will be outdated before you can fully adopt them.
- It is hard enough to keep pace with technological advancement in the age of AI, but if the vendor is not committed to innovation, your tools will be outdated before you can fully adopt them.
How do they measure success?
Do they help clients get value from the system? Most tech platform disappointments come down to poor adoption and low usage across the org. Any change to an established investment process is a big ask of the team. Ask what happens after go-live and who helps you get value as your process changes, who trains the analyst who joins next year, and whether the vendor has resources to help you succeed.
Can you ask an AI agent about your data?
This has recently become one of the more important questions as most funds look to AI to improve their investment workflows and operation as a whole. A brand new way of interacting with your portfolio data has opened up many new opportunities for investors.
An AI assistant that can read the public internet is useful, but an AI agent that can reason over your positions, understand your notes and your models, and answer in your context, can add much more value to the team. Every vendor has AI features now, so asking whether they have it tells you very little. Ask whether you can point an agent at your own data, get an answer, and see the citation showing where it came from.
If the answer arrives without a source you can check, it is not fully usable in an investment process. You need to trust the agent like you trust your data.
Our thoughts
We built EDS as the investment system of record because we think fragmented data is the problem, and no amount of point solutions fixes it. Research, portfolio construction, factor risk, attribution and governed AI belong in one cloud-native platform, with entitlements that let you decide who sees what and an MCP layer that lets AI reach your data with citations attached.
The test we would apply to any platform, including ours, is simple. Ask it a question only your own data can answer, and see whether it answers correctly.
See Fusion AI on your data: https://equitydatascience.com/fusion-ai/
