Catalyst CortexCatalyst Cortex

Your research, answering questions wherever your clients start

Your clients don't always come to you first. Your research can answer them in two places: your own platform, and the AI tools they already work in.

They do different jobs. We'll help you work out which one to start with.

catalystcortex-hero-image
  • 275000+ users

    on Content Catalyst platforms

  • 20+ years

    building for analyst research

Where you probably are

Most firms we speak to are in one of four situations. Each leads somewhere different.

cited-sources

Your clients have started asking

A subscriber wants to know whether your research can be reached inside their own AI tools. That conversation moves quickly once it starts, and your answer sets the direction.

Start with Gateway Catalyst
mycatalyst-hero-image

Your analysts are the bottleneck

The firm knows an enormous amount and finding it takes too long. Nothing client-facing has to happen for this to be worth fixing.

Start with MyCatalyst
consultants full

You've been asked to have a view

A board or an owner wants the firm's position on AI, and you would rather arrive with an answer than a defence.

Start with a conversation
testing-img

Something started and stalled

The connector worked. Then the licensing and permissions underneath it turned out to be the actual project.

Start with Gateway Catalyst

34%

of analyst research leaders expect their subscribers to want to pull research content into their own AI tools.

Source: Content Catalyst Analyst Research Leadership Survey.

Two doors onto the same research

MyCatalyst

AI chat inside your own platform

Your analysts, then your subscribers, ask your research a question in plain language. Answers come only from what you've published, with the source named and dated.

Explore MyCatalyst

Gateway Catalyst

Your research inside your subscribers' AI tools

A subscriber asks a question in their own AI assistant. The answer comes from your research, under their licence, with your name on it, and you can see that it happened.

Explore Gateway Catalyst

You don’t have to block AI, and you don’t have to bolt a chatbot onto the side of your portal.

Same content, same licences. What differs is where your subscriber is standing when they ask.

Behind both doors: Catalyst Cortex

Catalyst Cortex is the engine behind both. Two decades of working with research publishers, built into how the AI behaves: how entitlements are applied, how sources are credited, and what happens to your content afterwards.

It is why the same rules apply whichever door a question comes through: your entitlements, your attribution, your content.

Not sure which applies to you?

Let's have a conversation
catalyst cortex

The connection is the easy part

Getting research to answer a question inside an AI tool is straightforward. Any capable engineering team can do it.

The hard part is governance, and it’s where builds stall.

entitlements icon

Entitlements

Every person reaching only the research they're licensed to see, across many clients on different terms, as those terms change. Lock someone out of what they paid for and you get a support ticket. Do the opposite and you get a much worse call.

source icon

Attribution

Every answer traceable to the report it came from, with your name on it, rather than blended anonymously alongside two of your competitors'.

usage stats icon

Usage visibility

The moment your research is read in a tool you don't own, the signal that told you what clients valued disappears, unless you kept it deliberately. It is the same signal an account manager needs at renewal.

security icon

Security & compliance

Where the content sits, who can reach it, what is logged, what happens when a licence ends. Behind our answers sit policies and controls built over two decades of serving enterprise subscribers. A team building in-house meets the same review with nothing written down.

Does AI train on your research?

No.

Training absorbs content into a model, where it becomes part of what the model knows. Retrieval is different. A small, relevant passage of your research is fetched, used to answer one question, and then gone. Nothing is added to the model, and nothing surfaces for people who haven’t licensed it.

Your research stays yours, and you can see what was used.

Should you just build it?

Some firms should. The largest houses have the engineering depth to build this and keep maintaining it. If that is you, you probably aren’t reading this page.

For most, it is a decision to run a software team alongside a research business. One firm of thirty set out to build a connector, ended up with four developers, and called itself “accidentally a software company”. Meanwhile the research your clients pay for queues behind an engineering backlog.

We are not telling you not to build, only that it is worth deciding deliberately. Most firms arrive at it one hire at a time.

How we work

We have done this before, for this industry only.

Since 2003 we've put research publishers' content in front of their subscribers. Hundreds of thousands of people reach research through platforms we run.

We are in the room while it happens.

We work with a small number of research firms. Your questions reach someone who knows research publishing.

We will tell you when the answer is no.

Sometimes the honest advice is that your content isn't ready, or the demand isn't there yet. We'd rather say so than sell you a project that disappoints.

An analyst firm already using AI to deepen client relationships

Knowledge Capital Group named client retention as the biggest benefit they expect from their Content Catalyst AI agent, now live with more than 800 analyst relations professionals.

Knowledge Capital Group

The real differentiator isn't just the quality of our research, it's how easily clients can access and apply it.

Jeff Turner
Jeff Turner President and Partner Knowledge Capital Group

Common questions

What is the difference between MyCatalyst and Gateway Catalyst?

MyCatalyst puts AI chat inside your platform. Gateway Catalyst puts your research inside the AI tools your subscribers already use. Same content, same entitlements, different place.

Do we need both?

They do different jobs. Most firms start with one, depending on whether the pressure is coming from inside the firm or from your clients, and add the other when there is a reason to.

Can subscribers only see what they are licensed for?

Yes. Entitlements are enforced on every answer, through either door. Subscribers only get responses drawn from content they’re licensed to access. Your existing permissions carry across, with no separate rules to maintain.

We have heard about MCP. Is that what this is?

Gateway Catalyst uses the Model Context Protocol, the standard that lets AI tools connect to external content sources. The protocol is the plumbing. The commercial work is everything around it, from licensing to attribution.

Is our content ready for this?

Clean metadata. consistent taxonomy and well-structured reports are what let AI return accurate, attributable answers. Most libraries need some work here, so we start with a readiness assessment to show you exactly where you stand.

Start with a conversation, not a demo

Tell us where you are: what your clients are asking for, what state your content is in, what you've already tried, and what you don't want to take on. You'll leave knowing which route into AI makes sense for your firm, and in what order, whether or not you work with us.

Not ready to talk? Try Gateway Catalyst, our MCP connector, in an interactive walkthrough. See what your subscribers experience in their AI tools, and how you stay in control behind the scenes.