Your guide to Moonsift’s AI Shopping Co-pilot

Your guide to Moonsift’s AI Shopping Co-pilot

What is an AI shopping co-pilot?

This is one of the first, if not the first application of AI in this way.

Therefore we’re still in the process of defining exactly what an AI shopping co-pilot should be. And you are a key part of that process!

Because AI is able to ‘understand’ the context and characteristics of each product, it can make smarter choices on what it picks for you 🤞

What is the Moonsift’s AI capable of?

  1. It has a visual understanding of every product, e.g. it can see if a dress is below the knee or not

  2. It doesn’t just identify keywords, it understands the context of your search, e.g. *"a winter jacket that's suitable for office wear"*

  3. It can search across thousands of different retailers (big and small) and millions of products

  4. It can find you similar items to any item already in your collection e.g. a handbag you like, but at a lower price point

  5. You can ask it to add or remove specific attributes e.g. you like this style of sandal but want to remove ‘flat’ and add ‘wedges’.

How do I use the AI co-pilot?

The AI is designed to help you to sift through items across thousands of retailers while also honing in on specific attributes.

Whether you ask the co-pilot to help you find items similar to something in your collection or start your search from scratch, you’ll be able to dynamically refine the results by asking it to add or remove specific characteristics e.g. a colour, a brand, buttons, sleeves or a specific material.

1. Finding ‘similar’ items

Find similar items to something in your collection.

For example, maybe you want to find similar options at a lower price point.

Or perhaps you’ve found an item that’s almost what you want, but not quite.

Here’s a real example. Let’s say you want a winter hat like this, but without the pom-pom.

  1. Save the hat to Moonsift,

  2. Open the item in your collection and click ‘Find similar items’,

  3. Then in the ‘Remove attribute’ field type “pom pom”

  4. The AI will find you the most similar hats online without a pom pom

See the results for this hat with the ‘pom pom’ attribute removed.


2. Prompting with words only

Because AI understands ‘context’, you can make purpose-oriented searches like *"a winter jacket that's suitable for office wear"*

Whether you're looking for casual wear, business attire, or something in between, the AI should be able to understand the underlying themes and provide results accordingly.

  1. Start by describing what you are looking for (just like when searching on other platforms). This is your base ‘Prompt’ for the AI.

  2. You will be delivered an initial set of results to review

  3. Now help the AI hone in on what you’re looking for by adding and removing attributes

Help us improve it

Your feedback will go directly to the people working on this.

They are eager to learn:

  • What is it currently useful for (if anything)?

  • What could it be useful for with improvements?

  • What additional features would you like to see?

Whenever you see results that you don’t feel are relevant to your prompts, please click the thumbs down below the product image.

Sometimes you will be asked ‘Why’ a result was not relevant, any feedback helps.

What is an AI shopping co-pilot?

This is one of the first, if not the first application of AI in this way.

Therefore we’re still in the process of defining exactly what an AI shopping co-pilot should be. And you are a key part of that process!

Because AI is able to ‘understand’ the context and characteristics of each product, it can make smarter choices on what it picks for you 🤞

What is the Moonsift’s AI capable of?

  1. It has a visual understanding of every product, e.g. it can see if a dress is below the knee or not

  2. It doesn’t just identify keywords, it understands the context of your search, e.g. *"a winter jacket that's suitable for office wear"*

  3. It can search across thousands of different retailers (big and small) and millions of products

  4. It can find you similar items to any item already in your collection e.g. a handbag you like, but at a lower price point

  5. You can ask it to add or remove specific attributes e.g. you like this style of sandal but want to remove ‘flat’ and add ‘wedges’.

How do I use the AI co-pilot?

The AI is designed to help you to sift through items across thousands of retailers while also honing in on specific attributes.

Whether you ask the co-pilot to help you find items similar to something in your collection or start your search from scratch, you’ll be able to dynamically refine the results by asking it to add or remove specific characteristics e.g. a colour, a brand, buttons, sleeves or a specific material.

1. Finding ‘similar’ items

Find similar items to something in your collection.

For example, maybe you want to find similar options at a lower price point.

Or perhaps you’ve found an item that’s almost what you want, but not quite.

Here’s a real example. Let’s say you want a winter hat like this, but without the pom-pom.

  1. Save the hat to Moonsift,

  2. Open the item in your collection and click ‘Find similar items’,

  3. Then in the ‘Remove attribute’ field type “pom pom”

  4. The AI will find you the most similar hats online without a pom pom

See the results for this hat with the ‘pom pom’ attribute removed.


2. Prompting with words only

Because AI understands ‘context’, you can make purpose-oriented searches like *"a winter jacket that's suitable for office wear"*

Whether you're looking for casual wear, business attire, or something in between, the AI should be able to understand the underlying themes and provide results accordingly.

  1. Start by describing what you are looking for (just like when searching on other platforms). This is your base ‘Prompt’ for the AI.

  2. You will be delivered an initial set of results to review

  3. Now help the AI hone in on what you’re looking for by adding and removing attributes

Help us improve it

Your feedback will go directly to the people working on this.

They are eager to learn:

  • What is it currently useful for (if anything)?

  • What could it be useful for with improvements?

  • What additional features would you like to see?

Whenever you see results that you don’t feel are relevant to your prompts, please click the thumbs down below the product image.

Sometimes you will be asked ‘Why’ a result was not relevant, any feedback helps.