AI
This article is based on hands-on experience from real AI projects across different industries. It’s not a how-to guide, but insights we’ve gained by doing this work firsthand.
How Many People Will Use It?
This determines how much value the solution creates and how much investment is worth considering in the first place.
It doesn’t mean the number of your customers or the size of your company, it means how many people’s work the solution will actually affect. That’s what determines the level of the solution, not whether the company is small or large.
If five people use it, a well-written prompt and a brief explanation might be enough. Even at a multinational. If fifty people use it, it’s worth documenting and testing more seriously. If hundreds, that’s a project, you need a team, maintenance, support. And it matters whether only internal colleagues will use
Roughly speaking: a few people, a matter of hours or days; a few dozen, it takes weeks; hundreds, and you’re looking at a serious project.
Things You Might Not Be Accounting For
There are a few things that weren’t really issues in traditional software development but matter with AI.
One is cost. If you have a ChatGPT Plus subscription or use Copilot, usage is included in the monthly fee, you roughly know what to expect. But if you build your own solution, you typically pay per API call. Not necessarily a lot, but you need to factor it in, especially if many people use it or it works with large documents.
The other is the AI Act. In the EU, this legislation regulates AI use. No need to panic, for most internal use cases, compliance isn’t complicated. But you need to know about it, and it’s worth discussing with someone who understands it. Better to clarify upfront than deal with it later.
Experience Transfers; Solutions Rarely Do
You might think a solution that works for the HR team in Hungary at a multinational would be easily usable in Austria too. The answer, maybe. But at minimum it needs fine-tuning, and you’ll probably have to rewrite parts of it. It’s not just the language, there may be different processes, different regulations by country.
The specific solution, the prompts, the settings, the integrations, often can’t just be moved somewhere else. Not even within the same company. Some things might work regardless of department, like a simple daily task list or email categorization. But more typically, what works for the sales team needs to be adapted for HR. Different data, different processes, different expectations.
What does stay with you, though, is the way of thinking. How you approached it, what mistakes you made, what you learned. That’s worth a lot, and you can use it anytime.
A Few Concrete Examples
“We need to write proposals faster.” Start with a good prompt, understand, through a simpler example, how and how much GenAI can help, then share it with your team. You don’t need to build an app right away.
“HR keeps getting the same questions.” A chatbot might be the way to go, but first check whether the answers are actually written down anywhere. If not, start there.
“We want to digitise the company with AI.” That’s too big a bite. Break it into specific problems and start with one. If that works, move to the next. There might be an existing service for it, you don’t always need to build your own solution, but implementation can still be a big undertaking.
“I saw a great demo.” Okay, but what problem would it solve for you? If you have an answer, let’s talk. If not, let’s talk about the question.
Technology keeps evolving. What’s complicated today might be simple in six months. But the experience you gain now stays with you. If you’re unsure about something, ask. Talk to someone who’s done it before.
Our next blogpost: Rebuilding Lost Trust in GenAI – Part 3
If you’re ready to explore AI solutions for your team, our AI Enablement program can guide you from defining the right problems to testing practical applications. Book a short conversation with us and see where AI can make a real difference.
Contact our team: https://attrecto.com/contact
Our AI Enablement program is designed to help teams clarify their challenges and explore where AI can truly add value.
