Felix Heilingbrunner

AI output optimisation

By making it easier for users to skim for the core information they needed to make a decision through a new component, I increased setup rates in our creation and onboarding flows. Beforehand, the outputs were walls of text and user tests showed that they were quickly skipped and flows abandoned, tanking our setup rates.

{
  "title": "Statistics",
  "content": "12 orders",
  "description": "in the last 30 days",
  "icon": "TrendingUp",
  "isCollapsible": false
}

The component is manageable through JSON, which allows AI to fully control the output and information presented to our customers. Making it easy for users to skim for the core information they needed to take the next action.

Adoption across the company

3 different teams adopted the component to improve their setup rates and user experience. This increased usability scores, setup rates and received positive feedback from customers, like “easy to understand” and “clear to follow”.

AI creation

AI created starting points for the event creation, helping over the empty page problem.

Detail view sidepanel

Quick review of event summary through a sidepanel detail view.

In chat responses

Adaptable responses in chat to present information in an easily to scan format.

Mail templates

Aligning the design of the mail templates with the in product experience.

Buy vs. Build: How I convinced C-level to invest in UX

I managed the trade-off between buying a solution or building our front end in-house through usability tests and an Ai prototype in under 10 days.
I build a prototype with AI and compared it with the provider's experince through unmoderated usability tests with 40 users. The provider's product failed to meet our minimum usability threshold of 70, while my prototype scored far above the threshold. These tests proved that the investment in building the experience ourselves was worth it, which convinced our C-level.

53

Provider

87

Threshold

Adaptive AI input

To explore how AI inputs can be improved, I explored adaptive GUI elements to refine prompts and give live feedback to users, reassuring them that the AI understands their needs.

Objects of love

Over the past few years, I created illustrations of objects I love in Figma by hand.