RPA AI Recorder
A more natural way to create desktop automations. I worked with a team across Paris, Athens and Seattle to bring screen interaction, cursor movement and voice instructions into one recording experience.

Desktop flow creation can ask people to understand the automation tool before they can describe the task they want it to perform.
I reimagined the recorder and reconciled the classic recorder, element picker and task coach into a consistent interaction direction.
The experience combines actions on screen and spoken instructions to make desktop flow creation more approachable.
Follow the recording concept from explaining a task to reviewing the generated automation.
A focused recorder guides the person into the session before the desktop task begins.
The person moves between applications and explains what they are doing. Voice and on-screen actions provide context together.
The recording leads back into the automation designer, where the resulting actions can be reviewed and refined.
The design work connected the interaction model, visual consistency and an end-to-end demonstration.
Voice instructions complement what is visible on screen and where the cursor moves. The recorder keeps that explanation beside the applications being used.
I created an end-to-end video demonstration for Ignite, working with data science, PM and development partners to make the concept concrete.
The classic recorder, element picker and task coach had to feel like parts of one experience. I focused on consistency across the interaction and visual design.
The project explored how multiple input modes can lower the barrier to creating desktop flows.
People can combine a demonstration with spoken instructions instead of defining every automation action upfront.
UI interaction, cursor movement and voice are brought together in the desktop recording experience.
The work included an Ignite demonstration and was selected as a key feature for Microsoft Build.
