AI Butler for planing events

Momento Technologies


My Role
Product Designer
Team
Co-founder
Front-end engineer
Marketing manager
Tools
Figma
Claude
Timeline
2025 - 2026

Description

Taking Alfred, an AI-powered event coordination app, from idea to App Store launch by deciding what to build first and what to leave out.

Context

Alfred was a small team building an AI-powered event app for iOS and Android, and we needed one consistent look across both without slowing down. I defined the visual styles, wrote them up as primitive and semantic tokens in a doc, and iterated on the prompt until Claude, connected to Figma through MCP, could generate the system from it. The component list came from those tokens. We built on Tailwind so both platforms shared the same foundation, and documented everything in Figma. That base let us build more consistently and iterate faster without sacrificing quality.

Alfred app new event settings screen

Challenge

How might we help hosts spend less time coordinating guests and more time enjoying the event, with a first version a pre-seed team could ship?

Design Approach

100 Interviews

We set a goal to conduct 100 interviews to uncover the problem around friends spending time together and after conducting 16 we noticed a pattern. Each friend group had a “coordinator” - person who were spending more time navigating the get-together like juggling group chats, shared calendars and back-and-forth messages about availability and logistics.

Defining the vision

I started from high-level mapping of the flow to get the general idea of the app’s structure before moving into details. It helped the team to set priorities right and align a roadmap which we can rely on through the process.

Testing through TestFlight

We gathered early feedback from a group of beta testers through TestFlight, helping us test and iterate faster before the official app launch. Their feedback highlighted areas that needed a closer look.

Key features

Put AI at the core of guest communication

Hosts spend more time coordinating people than planning the actual event. We put an AI agent at the center handling outreach, availability gathering, and Q&A, shifting the value proposition from "create an event" to "host without the coordination overhead."

Alfred app screens with the AI agent handling guest communication

Unify planning, event details, and group conversation in one space

Hosts and guests were juggling calendars, group chats, and shared documents separately. By merging the itinerary, logistics, and conversation into one location, everyone stayed synchronized from planning through day-of, eliminating context-switching and coordination friction.

Alfred app screens with itinerary, logistics and group conversation

Share memories

After one of the usability testing we noticed that people were more excited about collecting and sharing memories after the event then simplifying the organization. As a result we created a sharing memory in a form of book, video or podcast.

Takeaway & reflection

During this project I got an opportunity of designing directly with a code base which was an exciting and wonderful experience thanks to Claude Code, I still used Figma to explore and brainstorm ideas. Though, the process of building features and apps is much shorter now, I realized how much more work we as a designers need to put into understanding who we’re building for, what problem we’re solving and why each of the feature needs to be built and each decision needs to be made.