Notes From the Field: How Investment Firms Are Applying AI to Proprietary Research

Over the past year — across VerityRMS customer advisory boards, roadmap sessions, and AI-focused discussions I’ve had with customers — one theme keeps coming up: firms are rethinking the role of their proprietary research.

I’ve heard multiple CTOs and heads of research say versions of the same thing:

“We know AI won’t replace our research process, but our research systems can’t limit us from staying ahead.”

The research systems built for the last decade weren’t designed with AI in mind. What worked then becomes a constraint now.

Most firms aren’t trying to become AI-first organizations overnight. But they are clear that their research environment needs to support the many AI paths they may take. And that their proprietary research and institutional IP are becoming far more valuable in that context.

What I’m Hearing in AI Conversations

AI Is Making Proprietary Research More Important, Not Less

As teams experiment with AI, one shift becomes obvious: AI is only as strong as the quality and context of the information it can learn from. And the only place that context reliably lives is inside a firm’s research corpus: their notes, models, meeting records, sentiment, internal viewpoints, and metadata.

When teams recognize that their research archive is an underused strategic asset, the research  management system shifts from a workflow utility to a foundational AI layer.

Every Firm’s AI Strategy Is Different — Adaptability Matters

Across conversations, a consistent set of needs shows up:

• Flexibility to support different models and LLMs.
• Ability to configure prompts around firm-specific workflows.
• APIs to support internal GPTs and firm-owned AI tools.

Any system that isn’t configurable quickly becomes a constraint. Teams want the freedom to apply AI in ways that match their process, not the vendor’s.

AI Is Killing Black Boxes & Raising Expectations for Integration

I participated in a panel at the Cutter Associates annual event in November, and one message was repeated throughout the discussions: black-box systems aren’t workable in an AI-driven environment.

Teams expect two-way openness, and things like MCP, APIs, and agentic workflows. They want insights generated by AI to feed into the research system. And they want the system to feed structured research context back into their internal models. If a vendor can’t integrate cleanly or walls off the data, firms notice immediately.

What We’re Seeing in VerityRMS

Customers Are Encouraging Us to Push Deeper Into AI

In the second half of this year, more than half of new VerityRMS implementations launched with AI enabled from day one, and among existing customers we’ve seen more than a 200% increase in usage of AI features.

The reason is straightforward: AI amplifies the value of proprietary research, and the research system is where that IP is structured.

Helping Teams Build the Story Behind Their Research

One of the most meaningful things I hear from customers is that AI is helping them build the story around their work in ways that were previously too time-consuming.

That includes:

• Understanding sentiment shifts.
• Connecting patterns across years of research.
• Identifying buried insights.
• Creating first-pass narratives grounded in metadata.

AI doesn’t replace judgment. It gives analysts more raw material and more clarity to build the story around their thesis. Plus, it can show them something they may otherwise have missed.

Making Interaction Natural

We’ve heard clearly that analysts shouldn’t need to leave their research environment to benefit from AI.

Features like AI Chat allow teams to ask natural-language questions of their own research — powered by their proprietary notes, tags, and metadata — and get immediate, cited answers inside the system.

It’s a simple idea: bring AI to the research, not the other way around.

Where We’re Headed

The future of AI in investing is still forming, but two priorities consistently show up across customer discussions.

Flexible, Model-Agnostic Support for Firm-Owned AI Strategies

What’s becoming clear is that firms are taking two equally valid paths: some want the research system as their primary AI interface, leveraging their chosen LLM through VerityRMS. Others want to power their own internal GPTs and models using our APIs and data sources as the foundation.

Both approaches require the same thing: the system must be open and flexible enough to support the firm’s chosen AI strategy rather than imposing one.

Related: How Firms Use the VerityRMS API

Surface What Deserves Attention So Teams Can Build a Better Story

As information volume and access grows, analysts don’t need more noise, they need help seeing what matters. This includes the ability to highlight what deserves attention, remind teams of risks previously identified, and surface shifts that matter to the portfolio.

For example, an alert might pop up saying: “There’s been a meaningful drop in sentiment across technology-sector meeting notes. You have significant exposure — consider re-engaging with the analyst team.”

Or: “You mentioned three months ago you wanted to revisit this name post-earnings. That’s now past, here’s what’s new with company ABC.”

The goal isn’t automation for its own sake. It’s enabling analysts to build a clearer, more complete story around what’s happening in their coverage and act on it sooner.

Bottom Line

AI isn’t replacing the investment process. It’s revealing where the process needs to evolve — and exposing the importance of proprietary research and institutional knowledge.

In that world, the research system becomes the connective layer between a firm’s research, its people, and its AI strategy. Our job is to make sure it stays open, flexible, and built to support the way fundamental investors actually work.

Is It Time to Upgrade Your Research Management?

Find, share, and act on insights faster — without losing control of your process. With VerityRMS, you can connect investment teams to the research, data, and tools they need while cutting out the busywork that slows them down.

Watch an RMS Walkthrough Now >>