When your AI learns how you work, you both get better

Project Catalyst | Part 3 of 4

Spend more time deciding and less time hunting across screens.

Getting ready used to mean hunting across your apps to form a cohesive picture.

For a lot of people in capital markets it still does, and AI has added a step. In a survey of 1,000 professionals in finance, healthcare and the public sector that HERE ran this summer, three in 10 of the AI users said they spend at least half of a typical workday re-entering information between AI tools and other systems.

Project Catalyst lets your AI put that picture together for you from your vendors' own screens, each opened with the answer already on it. Once the AI can bring up those screens, it sees the context of the work before, during and after it's brought in to help. As you teach it how you work, it can use the Catalyst MCP (Model Context Protocol, the standard for letting AI reach into software) to bring the whole workflow back for you, set up the way you want it, so you start from the picture instead of the hunt.

The AI sees what it needs to see

Catalyst lets your AI meet you where you are. It understands what you're looking at and what you're trying to do, so you don't have to explain your situation or paste it in before it can help.

Your AI knows what each of your apps is for, so it brings up the right screens, even for a question it has never seen before.

From there it works alongside you like a colleague who shares your context, only faster and across more of your applications at once. It connects what your firm knows and brings it to the point of action, where the decision gets made.

Catalyst at work on a new order

A new order lands, and you ask your AI to get you up to speed on it. It finds the order and builds a dashboard made up of your own apps, each already showing the stock. Here it's brought in why the stock is moving, what it'll cost to trade, who's best to trade it with, and where the price has been, all of your internal and third-party intelligence in one place for you and your AI.

Then it drafts the note to your portfolio manager and waits for your approval. You're both looking at the same screens, so you can test its read against the source and push it further before anything goes out.

A portfolio manager preparing for an investment committee, or an operations team checking a corporate action against open orders, has the same pieces to pull together.

When the screen you need doesn't exist yet

Sometimes your firm builds the missing screen itself, and your AI can do that for you in HERE Studio, the app builder inside the Catalyst platform. It works from the APIs and MCP connections you already have. Sometimes the vendor builds it, the way viaNexus built its own apps in Studio and BTON built a widget that runs only in Catalyst. Either way, it sits in the dashboard beside the rest of your screens and stays there for next time.

Studio fills the gaps between the apps you already rely on.

Once it sees the workflow, it can keep it

When the dashboard looks right, you can tell your AI to save it. Because Catalyst has hooked it into your workflow, your AI can remember which apps you used and how you like them laid out, along with the calls you made along the way. We call that workflow memory.

The next time an order lands, one phrase is all it takes, and the dashboard comes up already showing the new stock. You decide what it keeps.
Your AI already remembers what you told it. What it misses is the work that happens across your apps, and that's what Catalyst lets it see and keep, whichever AI you use.

Your AI keeps pace with your thinking

The call stays yours, and the point is to make it with a clearer head. While you form the view, your AI keeps the work moving around you, pulling up what you ask for, drafting what you'll want to send and remembering how you got there. Each round between you sharpens your read, and you get to a better decision faster.

One benefit comes free. When the way your best people get ready is saved, it can become where everyone else starts, including the analyst who joined last month.

What runs today?

Catalyst already works with about a dozen integration partners, from order management to market data and pre-trade analytics, and the architecture is in place to make adding the next 100 straightforward. Our latest video shows your AI building a dashboard from five of them around a new order.

A pilot runs on the applications you already have, under your own logins and permissions. Your AI sees what you're cleared to see, and anything it saves stays inside your own environment.

Each dashboard arranges itself around who you are and what you need, because the AI knows what each screen looks like and how best to present what's on it. We're working with the software providers you already rely on to make this the new normal, which is the role HERE has played on the capital markets desktop for more than fifteen years, first as OpenFin.

AI can be pointed at almost anything, which is one reason so many pilots spread thin and stall. We'd rather start with one workflow your team runs every day, so the difference shows up in work you'd be doing anyway.

Get involved

We're running pilots now, inside firms' own environments, on their own models. If you want to see it on your stack, or you're building something a firm's AI should be able to reach, talk to us.

Details are live at here.io/project-catalyst. What gets built first is decided by the firms we build it with. Email us at catalyst@here.io.