At the earliest stages of a startup, I love to focus the team on business objectives with a Business Model Canvas. This tool helps identify value propositions and product-market fit. A side-effect is it develops language to describe the product (for pitch-decks).
To get things started, I worked out some high-level steps for the user to get onboard and find an assistant to complete tasks. The concept included an AI bot that would identify the user's needs and find them an appropriate assistant.
Through iteration, we determined that the bot would be best implemented in an message-based interface. I developed a welcome sequence with a questionnaire to handle the bot's 'interview'.
A few low-fidelity wireframes helped determine the direction for the app's second function - task status tracking. We played with a few layouts including a card-based interface.
An interactive prototype and some animation work in Principle helped visualize the interaction of the message centre and the task tracking.
Mockups and animations for product presentations provided enough excitement to get several developers and investors interested in the startup. Final designs of the published app differ in colour and layout but this early UX and UI DNA still shows in the final product.