The product: find the right help for an everyday detail
What color is a shirt? What does a package say? These small questions sometimes need another person's eyes. Be My Eyes brings that help to a phone: Be My AI describes a photo in text, while volunteers see the user's surroundings through video and talk with them.
The two approaches serve different needs. A still photo can provide an initial understanding of a scene; a live conversation is more useful when someone needs to move an object while checking its details. Users choose where to start. They do not have to ask AI first before contacting a person. The product and its support options
This article uses identifying clothing on a phone as its example. The app provides free support on iOS and Android, and registered users must be at least 18. Set a preferred language when signing in and allow camera, microphone and other permissions as needed for the chosen form of help.
The core workflow: get a description, then choose a follow-up
Suppose you want to distinguish the colors and patterns of two tops. Place them in suitable light and keep unrelated personal materials out of the photo.
1. Photograph the objects you want to understand
Open Be My AI, choose Take picture, and photograph the clothes with the rear camera. Wait for the description. An internet connection is required, and an unstable connection can delay the reply. If the image is unclear, taking another photo is more useful than repeatedly asking about the same blurry one. The camera entry point Connection and photo requirements
2. Ask about the missing detail
Choose Ask more to add text or another photo. For example: “Is the top on the left plain or striped?” If the clothes overlap, separate them before adding a photo. The aim is to supply material that can answer the question, rather than simply asking for greater accuracy. Be My AI does not currently accept video. Adding photos and questions
3. Contact a person when you need live clarification
If questions remain, choose Call a Sighted Volunteer within Ask more. You can also leave the photo conversation and call from Get support at the bottom. A connected call uses one-way video and two-way audio: the volunteer sees your camera feed, and you can talk to each other. Start with a specific request, such as “I want to check which top has stripes,” then adjust the camera as the conversation requires. Ways to contact a volunteer
In mainland China, volunteer calls are unavailable from 23:00 to 06:00 the next day; Be My AI remains available 24 hours a day. If you need a person to confirm something and cannot call now, try later or ask someone you trust. An AI description does not mean a person has confirmed it. Regional and time restrictions
4. Respond to connection feedback
Matching prioritizes your preferred language. If no one connects after one minute, the official advice is to end the call and try again. If you accidentally move the app to the background during a call, the volunteer may lose the camera view. Return to Be My Eyes to restore it; you do not need to redial for that reason. Connecting and restoring the picture
The intended result is to understand the clothing differences that matter to you. For sensitive personal or financial material, the official guidance is to ask trusted family or friends rather than give it to Be My AI.
Design lessons: let users decide how to continue
Two channels serve different information needs
A photo description turns a scene into material for follow-up questions; a video call lets two people discuss a changing situation. These are not simply higher and lower levels of the same service. The useful design lesson is to understand whether a user needs a static explanation or live collaboration, then make that choice easy to find. The tradeoff is that users must judge when to switch approaches, so each entry should explain the help it provides.
A follow-up can add evidence, not just different wording
When a pattern is hidden, another question cannot recover what the photo never captured. Allowing a new photo in the conversation gives the user a way to improve the input. Similar products can make both “What material is missing?” and “What do I want to ask?” actionable, instead of trapping users with the same input.
Put the next route where the question arises
The human-support option appears among the photo conversation's follow-up actions, so users need not work out which settings menu might contain help. Call availability and connection waits also expose a tradeoff: a visible entry does not mean an immediate connection at any time. An assistance flow should make the next action easy to find and explain what is currently available.