Why I Built a Free Responsible AI Gap Analysis for FileMaker

The question I keep getting from clients is some version of the same one. Can AI do this? Can we just use AI for that?

And as much as I want to say yes, because I know how powerful these tools are and how much they can actually do, the question I want to get to is a different one: how do we use AI the right way?

That’s the work of Violet Beacon, which is helping companies and individuals use AI the right way. The Responsible AI Gap Analysis for FileMaker is one piece of that: a couple dozen plain-language questions, about fifteen minutes, built for any company using FileMaker that’s getting these same questions and doesn’t have anywhere clean to point. You take it, you schedule a call, a consultant reviews your responses, and you get your results walked through in context.

Who this is actually for

If your company runs on FileMaker, whether you have an internal developer, a small team, or a vendor maintaining your solution, and you’ve started adding AI features or thinking about it, this is for you. Same goes if you’re a Claris partner shop building for clients and getting these questions from them. The questions are written for small and mid-sized organizations that don’t have a dedicated compliance function, regardless of whether the AI is being built in-house or by someone you’ve hired.

What I keep seeing in client work

Across the FileMaker teams I’ve worked with this year, both in-house developers and partner shops, four things come up over and over.

Transparency. Teams have an AI feature in production and can’t fully explain how it’s making decisions. Everyone meant well. It just never got written down until a customer asked.

Data handling. It’s often unclear what gets sent to a cloud model, what stays local, and what the client agreed to when they signed the proposal a year ago.

Disclosure. Some AI features are obvious. Some aren’t. The ones that aren’t are where the awkward conversations happen later.

Documentation. Most of what teams know about how their AI works lives in someone’s head. That works fine until that person is on vacation, or a client asks for an audit trail.

What the Gap Analysis covers

It’s a couple dozen questions and takes about fifteen minutes. Which ones you see depends on where you already are, so a team running AI in production today gets a different path than a team still deciding.

The questions walk through the things that actually decide whether your AI use holds up: who’s formally responsible for AI decisions in your environment, whether there’s a written policy for using AI tools with company data, how you decide a new use case is appropriate, which of your data is sensitive and what actually leaves your system when FileMaker’s AI script steps run, what happens to AI output before someone acts on it, whether AI call logging is turned on, how you picked your provider, and whether leadership understands where these features fall short, all in plain language.

You take it. You schedule a call as part of submitting it. Your results arrive at that call, after a consultant has reviewed and audited your responses. That review step is the whole point: context is what turns a list of gaps into something you can act on.

What happens after

When you submit the assessment, you schedule a call with a consultant as part of that submission. Between submission and the call, the consultant reviews and audits your responses. Results are delivered during the call, walked through in context.

I run that call as education. Most teams taking this assessment are doing it because they want to understand what responsible AI actually looks like in practice, and a raw report doesn’t get you there. What I want out of that conversation is for you to leave with a clearer sense of where your current AI use is in good shape, where it isn’t, and what a reasonable next step would look like for a team your size. Sometimes that next step is something we’d do together. Often it isn’t. Either way, the temperature read is the value.

Why it’s free

There are three honest reasons.

One, the people who need this most are often the people who can’t afford a full governance engagement. If responsible AI in small businesses depends on people being able to pay for it, it won’t happen.

Two, the conversations I have with people about this teach me more about the state of responsible AI in the FileMaker community than any single client engagement does. That helps me build better tools and write more useful things.

Three, this is the kind of thing I wish someone had handed me three years ago.

Back to the question

The clients asking me can we just use AI for this are right to ask. They’re paying attention. They can see what’s possible and they want in.

The Gap Analysis is how you answer that question yourself. It shows you where you can say yes confidently, where you need to do some work first, and where the right answer is “not yet.”

That’s the version of responsible AI I care about: doing the work in a way that holds up.

Take the Gap Analysis.

I’m Kate Waldhauser, ISO/IEC 42001 Lead Implementer, credentialed in the NIST AI Risk Management Framework, and a Certified Claris Partner. I run Violet Beacon in San Marcos, Texas.

If you take the assessment and something surprises you, I’d like to hear about it.