Find a bounded task within design experience
In an April 2026 Indie Hackers interview, Fernando Pessagno described his design background and tools including aiCarousels. The official site identifies him as founder and presents AI-assisted writing, adaptive text and preset colors and fonts for social-content creation.
This is public-source research without a subscription test. The site's one-person-team description is a personal account, not an independently verified list of employees, contractors and partners. Interview revenue and growth figures are unaudited and are not used to prove results here.
Users buy a route to completing a task
The following is editorial analysis. A professional designer may want complete control, while a store worker may simply need information arranged across several images. Both are valid needs, but a small first version may not serve both. Identify the user and where the task stalls.
Imagine a neighborhood repair shop explaining three services. Its manager has service facts and photos but needs help ordering them. Observe whether they can define each page's topic, detect excess text, replace unsuitable images and export usable files. This is our example, not an actual aiCarousels customer.
Fewer choices require more understanding
Preset typography, colors and layouts do not eliminate design decisions; creators make many earlier. Record choices that repeatedly cause hesitation and adjustments users must retain. Brand names must remain complete, opening hours must not be shortened away and service conditions must not vanish for neatness.
AI can propose wording, while users check facts and tone. If text is shortened automatically, show what was removed rather than only the prettier page. Making one revision easier may support continued use more than adding generation options. These are design ideas, not verified features of the source product.
Check whether free entry points serve the same people
Fernando also discussed connections between free tools and other products. A question for founders to test is whether free-tool visitors next need a task connected to the paid product. Traffic and tool count do not answer that question.
An image-size checker may attract people preparing to publish; an unrelated entertaining quiz may not, despite wide sharing. Without unnecessary personal-data collection, observe completed tasks, voluntary editor entry and return visits, and ask why people leave. Record free and paid behavior separately rather than converting the former into certain orders.
Before expanding, check that the original task remains clear
Requests for long articles, video or managed accounts require reassessing who asks, which experience transfers and whether the service changes. For teams of one to ten people, another business line means additional support, assets and expectations. Partner for capabilities or defer them according to evidence of demand.
This case concerns design experience entering AI tools, not a ten-day success timetable. Domestic projects need outputs adapted to actual platforms, content and customers. Sustainable business requires usage, cost and feedback evidence; an overseas founder story does not establish local policy eligibility.
Further reading: users versus paying customers; reviewing content before publication; selecting a first collaborator.