Vestmark, Inc. Innovation, Technology & Agility

What is the unique story that you feel your company has with AI? If you were writing about it, what would the title of your blog be?

I think the title would be something like “AI — it’s not about you” or something to that effect. The reason being, there is a tremendous focus right now on how AI will drive time savings for financial advisors (or really, anyone leveraging AI) so they can have more time to do their other, non-AI impacted parts of their jobs (i.e. meet with their clients). But I think the real power of AI comes from the kinds of services and capabilities that AI unlocks as a result of how AI saves us time. Case in point: if we can dramatically reduce the time it takes to generate a bespoke investment proposal — using an AI agent that leverages existing Vestmark software and inputs from the advisor — we also unlock entirely new levels of flexibility for running “what-if” style analyses on those proposals. So yes, we have saved time in the creation of one individual proposal but the same thing that saves us time also unlocks a whole new kind of service — and that is the real power of AI.

What was a monumental moment for your team when it comes to your work with AI? 

I don’t think I’ll ever forget the first time we asked our very basic agent to update a client’s email address and saw it reflected in our platform. Just that very simple proof of concept made the possibilities of Agentic AI feel real. Because if we could get an LLM to take a totally unstructured query, find the right set of APIs to execute the task, AND have it reflect in the platform for a user to see and validate, you have all the pieces of the puzzle for a full-blown AI-based assistant. It was that simple proof of concept that we used internally to show people the promise of agentic AI to build additional momentum behind expanding the capabilities out to where it is today.

What challenges did your team overcome in AI adoption? 

I think the biggest challenge in AI adoption is having the discipline to know when AI is a good fit to solve a real problem… and when it isn’t (or, isn’t quite yet). It can be very tempting to approach AI as a solution or replacement for whole swaths of human time and effort, but you need to take a step back and appreciate where humans add real tangible value above just their cost or ability to push keys on a keyboard. For example, in wealth management, many of our users, financial advisors, carry a real fiduciary duty to their clients — which means they are going to be held accountable for decision making around which products a client ultimately invests in. 

They need to be in control throughout that process because that is the agreement they have with their end client. If as a software platform we were to take the financial advisor out of the loop at critical points in the portfolio management process — even if it “saves time” — it misses the whole point of our platform to our user; to help them fulfill their objectives in serving their clients. And that comes back to a major sticking point I think many firms are going to face in driving AI adoption — building trust.

What types of products or services does your engineering team create? What problem are you solving for customers?

At Vestmark, we provide a wealth management platform that enables our customers to efficiently manage and grow investor portfolios. Our solution helps our clients streamline operations and focus on delivering value to their investors.

My team is responsible for the ingestion and analysis of financial records from partners each day, ensuring the data is accurate, reconciled and ready to trade. There are many moving parts to be orchestrated, executed and monitored to maintain the highest data quality. A single client environment can have hundreds of tasks running daily to keep the system operating smoothly.

By automating and monitoring these complex processes, we help our clients focus on decision-making and investor outcomes instead of operational overhead. This transforms a previously time-consuming back-office challenge into a reliable, scalable and transparent workflow.

Tell us about a recent project where your team used AI as a tool. What was it meant to accomplish? How did you use AI to assist?

With hundreds of daily tasks for each client, it was difficult to visualize progress or identify areas for improvement. Using the AI tool Cursor, I built a script that automatically generates a timeline image of each process. Previously, a timeline could be created manually in Excel — a time-consuming effort. With the script, a simple query now produces the image, saving hours and making the data easy for anyone to understand. Cursor made development fast; what could have taken weeks was completed in an afternoon.

The script worked well but still required manual queries and lacked interactivity. Later, using prompts with Claude, I created an enhanced web app in about 16 hours. The web app live-queries data and lets users drill down from a summary chart all the way to individual tasks. It provides clarity and insights that were previously unavailable, turning a complex process into a transparent, data-driven view.

What would that project have looked like if you didn’t have AI as a tool to use? 

When I asked Claude to estimate the development timeline for the web app we had built, it produced a six-month project plan for a full team. We built it in a week.

Without Claude, this project would likely still be on the long list of great ideas we don’t have time to develop. With AI, we delivered immediate value and turned an idea into a working prototype in days.

Debugging is another area where Claude shines. Normally, a single issue can consume hours of trial and error. Claude tirelessly tests different approaches until it finds a solution, making it faster and easier to work through problems.

Looking ahead, I see that AI will enhance our ability to add new features to our product. These tools can feel almost magical in how quickly they deliver results, but they don’t replace solid design and architecture. Success still depends on clear requirements, sound data models and scalable designs. AI simply shortens the path from concept to reality.