> ## Content Index > Fetch the complete content index at: https://intelligence.artificialrealitymedia.com/llms.txt > Use this file to discover other available public pages before exploring further. # Five hundred loops: what building your own AI tool actually took - URL: https://intelligence.artificialrealitymedia.com/five-hundred-loops-building-your-own-ai-tool/ - Published: 2026-09-02T12:12:00.000Z - Updated: 2026-09-02T12:12:00.000Z - Description: Jennifer Kerum tried to build her own client intake platform with no technical background. Her verdict, and the six checks to run before you start. - Author: Galina Fendikevich - Tags: Yacht charter, #For AI Buyers, #What's not working, #Import 2026-09-03 13:11 Building your own AI tool without a developer is possible and still costs more than the marketing suggests. Jennifer Kerum, co-founder of Anchor Croatia, a luxury yacht charter brokerage, tried it for her own platform and website, and her summary is that she should not have had to *"go through like five hundred loops."* No-code means no programming language. It does not mean no learning curve. She is not a reluctant buyer. She is an operator actively trying to build, with a specific application in mind, who cannot find something usable. ## Why is a yacht broker building her own software? Because nobody is building it for her industry. Kerum's assessment of charter, given to Artificial Reality, the AI buyer-evidence podcast hosted by Galina Fendikevich: > "You do have some software that is catered to charter companies, but I feel like in general, charter companies and even sales companies and yachting almost get forgotten in the AI world. There isn't a lot of tools that cater to us. And it would just be amazing to have more tools available." That is the calculation behind most operator builds. Where a vertical product exists, buying is almost always cheaper. Where one does not, waiting is not a strategy. ## What does the wall actually look like? It looks like a build that runs, and a gap between running and finished that nobody priced. In her words while it was happening: > "I would really love the accessibility and easy integration. Right now I am trying to build this platform and this website, and I'm really struggling to find the tools to actually help me get to where I am. We're doing it. It's just like — it's like this, you know?" Her benchmark for what accessible should mean is precise and fair, and it is not developer tooling: > "When you think about when you used to build a website twenty years ago, you needed to have a web developer, right? And now you have so many different programs that even if you didn't know anything about websites, you could just plug and play." That transition — from specialist skill to consumer product — has happened for websites, storefronts and payments. It has not happened for AI applications. Which is why the honest verdict is hers: *"It would be nice to see it more accessible, and to not have to go through like five hundred loops."* ## What is the cost nobody quotes you? Training your team, as a separate second project after the software works. > "Even once we implement all of this, then we're going to have to go to a training program to actually train our employees. So it's going to be a two-step process. But it's also exciting, you know." That second step appears in no vendor's pricing. Budget for it before you start, and ask any vendor you shortlist what onboarding a non-technical team actually involves in practice. ## What should you do before starting a build? Six checks, in this order. They are drawn from what went wrong here, and each one is cheap compared with discovering it three months in. - **Split the problem before shopping.** The customer-facing conversation and the internal workflow behind it are usually different tools. Deciding which you are solving first stops you evaluating products against two unrelated requirements. - **Build one narrow thing in a trial week.** Not the whole platform — one flow. A tool that cannot get you to a single working result in a week will not get you to a platform in a quarter. - **Check the deployment path early.** Confirm the output can be embedded in your website, on your hosting platform, before you evaluate anything else. This is the most common late-stage disappointment. - **Ask what happens when it is wrong.** For an agent quoting prices or availability to clients, ask specifically how answers are constrained to verified data, and what it says when it does not know. Confidently wrong AI has already cost this operator. - **Budget the second step.** Implementation, then training. Two projects, not one. - **Confirm you can maintain it yourself.** If changing a question in the intake flow requires the person who built it, you have created a dependency rather than a tool. ## Is the gap you are hitting your own? No. Kerum's description of how AI capability is actually distributed is the most quotable line in the interview, and it is about the interface, not about you: > "I feel like there's only a small portion of people that really understand how to use AI. And then I feel like there's everyone else that is just like: I'm hungry, what do I eat?" ## AI tools named in this report **None.** Jennifer Kerum named no product she used or abandoned in this attempt, which is itself the finding: in yacht charter there was nothing obvious to name. Artificial Reality found no vertical AI builder for charter businesses; the horizontal no-code options are covered separately. Inclusion in that coverage is reporting, not endorsement. If you are starting this month, pick the single flow that costs you the most time today, give it one week, and stop if you cannot get one working result. That week is the cheapest way to find out whether the tools have reached your industry yet. ## Where this comes from S1E2: How a luxury yacht company uses AI without losing the personal touch — the full interview with Jennifer Kerum. Listen or watch: [YouTube](https://www.youtube.com/watch?v=i3jt4fs02FM&ref=intelligence.artificialrealitymedia.com), [Spotify](https://open.spotify.com/episode/5Qk0cYaTkTmeLAYkXsmBxN?ref=intelligence.artificialrealitymedia.com) or [Apple Podcasts](https://podcasts.apple.com/us/podcast/how-a-luxury-yacht-company-is-using-ai-without/id6780699025?i=1000773179066&ref=intelligence.artificialrealitymedia.com).