The old impactmind.ai described an “AI platform,” a phrase broad enough to hide the decisions that matter: which workflow, which users, and what the product should help them do.
We rebuilt the site because our product story had become more specific.
A portfolio needs an order
ImpactChat is the first product visitors meet. ImpactRole follows as a second available product. ImpactBrain, ImpactExam, and ImpactAgent are presented as early-access products, with their status kept beside their names.
That order matters. A portfolio page can make every product look equally mature even when it is not. We wanted the website to distinguish what can be discussed in a demo today from what still belongs in a scoped pilot.
Interface previews need boundaries
We did not have a complete set of current product screenshots for every page. Instead of disguising concept interfaces as production evidence, we label them as interface previews.
The previews show the workflow we are designing around. They do not prove a customer deployment, a security certification, a performance metric, or a connector that has not been confirmed.
The rules we kept
- No fake proof. No placeholder customers, invented usage numbers, or decorative compliance badges.
- Status stays visible. Early-access products should not look quietly production-ready.
- AI is not the outcome. Product copy starts with the work people are trying to complete.
- One source of truth. Product names, routes, statuses, and demo interests come from the same registry.
What the website is for
The website should help a visitor understand the product portfolio, choose the relevant conversation, and decide whether to open the meeting calendar. It should also make declining tracking possible: HubSpot chat and the meeting embed load only after explicit action.
That is a narrower job than telling the entire company story at once. The narrower job produced the better website.