Design an AI-based application in a store that would help customers check out clothes and try them on.

Clarifications:

1. Is it a store that carries clothes alone?  Or even other merchandise? Like jewellery, hats, shoes, watches etc?

2. Is this targeted for a certain group of users:

• Shoppers with a very high sense of fashion and brands and spend a lot on clothes
•  Shoppers who struggle to make a decision of what they like and what they don't
•  Shoppers who find it really hard to find clothes that suit them or fit them?

3.  Does the app run on a kiosk or is it a mobile app ?

Assuming the interviewer says:

Clothes that carry clothes and merchandise.  

For Shoppers who struggle to make a decision of what they like and what they don't.  

App runs on mobile phones

Let me come up with a few pain points/use cases we are trying to solve for the user group:

1.  Indecisive shoppers often don't remember what they already have and end up trying to decide between buying stuff they already have and don't.  They don't come with a plan for shopping.  

2. Shoppers need a second person like a close family or a friend to tell them what suits them what doesn't and also add convincing comments like "Hey you already have that color.  Or that color didn't go well with you the last time you wore it.  Go for a different color or a lighter color"

3.  When shoppers come with a fixed budget and end up liking more items and would like to purchase they are unable to decide which ones to pick and which ones to drop.

4. Shoppers tend to miss out on buying matching merchandise and end up making a bad decision of buying other clothes that they already have instead of thinking of adding matching merchandise to the clothes they bought that adds to their style and also keeps them from buying the same clothes they already have.

Now let us go through the user's workflow for the app:

User walks into the store ==> App should recognize which store it is based on the location in the mall or a separate store by itself ==> 

App should be able to :

1. Tap into the Store's website and have access to all the clothes and merchandise the store 

2. Have access to the pictures of the user from the user's social media platforms  

3. Maintain the user's history of frequently looked items in that store and also other related stores

4. Maintain some models and combinations of popular styles of clothes and merchandise.

==> App should be able to gather data from the above 4 sources and suggest the user based on his/her tastes the clothes available and also be able to suggest good combinations of clothes and merchandise.  Saves a lot of time and effort for the user and indirectly encourages the user to buy more as the customers don't give up spending so much time trying to think what they like but instead the app suggests them what to buy.

Solutions:

1.  Have the App detect the store based on the location

2.  Have the app connect to the store's website, user's social media pictures and the pictures on the phone.

3. App should be able to remember what the user looked the most in that store and related stores.

4. App should be able to maintain a library of some popular trends, combinations and merchandise

5. The core algorithm should be able to suggest all the available clothes and merchandise based on the data collected in the above solutuons 1,2,3 and 4

6. App should also be able to suggest nearby stores that have the items that the user might be looking for.

7. App should be able to provide updates to the user about an item that becomes available in a store.

8.  App should also provide a way to shop online from any selected store if the item becomes available in a store the user visited before and was not able to find it.

Prioritization (MVP):

1. Have the App detect the store based on location

2. Have the app connect to store's website , user's social media pictures and the pics on the phone

3. App should be able to maintain a library of some popular trends, combinations and merchandise

4. Should be able to suggest all the available clothes and merchandise based on the data collected in the above solutuons 1,2,3 and 4

Backlog:

5. App should be able to remember what the user looked the most in that store and related stores.

6. App should also be able to suggest nearby stores that have the items that the user might be looking for.

7. App should be able to provide updates to the user about an item that becomes available in a store.

8.  App should also provide a way to shop online from any selected store if the item becomes available in a store the user visited before and was not able to find it.

Metrics:

1. DAU, MAU

2. Customer acquisition : % increase of new user signups in a week. 

3.  Survey of customet satisfaction after the cu uses the app and shops and steps out of the store

4.  CTR of suggested combinations to view in a larger view

5. Avg Time spent on the app while in the store.

6. Avg time spent in the store

Tradeoffs:

1. This will not appeal to customers who are very conservative and simple when it comes to buying clothes

2.  Not all stores might have online websites, so in that case the app can only suggest based on the user's photos, tastes and popular trends but not ppint out whats available in the store.