“MenuMate is an OpenAI-powered web app that aims to reduce food and social anxiety for People with Blindness or Low Vision”
Project Snapshot
Menu-Mate transforms inaccessible restaurant menus into screen-friendly formats, enabling Blind and Low Vision users to access and interact with the menu. Through a chat feature, users can obtain detailed and accurate descriptions of menu items, helping them avoid unpleasant dining experiences. During this time, I was responsible for:
1:1 in-depth interviews
The product was informed through insights gathered from 3 Blind and Low-Vision users.
Data Analysis
Transcripts from interviews and usability tests were analysed to look for product opportunities and gather evidence of efficacy
Usability Tests
The final product was tested with 2 Blind and Low-Vision users in 60 minutes sessions
Foundational Research
Does this ever happen to you? You order a burger thinking that it will fit perfectly in your hands only to be unpleasantly surprised by its towering size?
Through two in-depth 1 hour long interviews that aimed to understand what information people from Blind and Low Vision communities need to make more informed food-ordering decisions, I learned..
Social + Food Anxiety
People from the Blind and Low Vision Community may experience social and food anxiety due to such unpleasant food experiences and increase their dependency on others to make food ordering decisions.
How might we help Blind and Low Vision users manage social and food anxiety from unexpected experiences by mirroring conversations with restaurant staff to provide accurate menu descriptions?
Presenting MenuMate!
Upload Users can upload an image of an inaccessible menu and then…
Load Transcript
Menu-mate converts it into a screenreader friendly format and finally..
Chat!
Allows users to ask questions from the menu to get more accurate info.
Usability Testing
Set-up
Users were presented with two menus in two study phases. First, they interacted with Menu A using their existing technology, and then with Menu B using our chat interface. Throughout, they voiced their thoughts and imagined ordering from a real restaurant.
Time: 75 minutes (15 minutes of pre-study and post-study interview questions)
Tasks
1) Can you decide an item to order from each section of the menu?
2) Consider one of the items from the menu and describe the following:
3) Portion Size b) Calories c) Flavor / Spice level d) Presentation e) Prize f) Potential allergens
Evidence of Efficacy
The usability tests were recorded, and the transcripts were analyzed to compare participants' responses when using their screen readers or other technology to access menu information V/S their responses with MenuMate. Users received more descriptive responses with MenuMate. The AI can analyze images of the items to provide detailed information on various attributes of the food, such as ingredients, meal size, texture, and more.
