With a bit of an understanding of your clustering goals and some sample data, we can easily get you comfortable with the quality of our results.

ANYTHING!

SimilarParts.ai only requires industry standard Comma Separated Value (CSV) Files as inputs.

Spreadsheet Software (e.g. Excel, Google Sheets, Open Office, or Apple Numbers) or Database tools (e.g. Microsoft Access, SQL Server, MySQL or Oracle) can ALL export a "CSV" File.

Typically, it is as easy as a "Save As" and then selecting CSV as the File Format... or an "Export" to CSV File.

Once you have this... SimilarParts.ai can easily import and process the data with ease.

A Comma Separated Value ("CSV") File is basically a Text File of your "table" data.

The First Row is called the "Header" and contains the column names.

Row numbers two (and beyond) contain your items and their corresponding values to the column names.

Each Row within the CSV File (including the Header) should have values separated by commas (","). This is the separator that helps define where values start and stop for each row based on the column defined in the Header Row.

This format is standardly exported by Spreadsheets and Databases, so you should not need to alter the format of the file.

NOTE: SimilarParts.ai DOES support multiple values in a single column - they are often called "Lists" or "Arrays" of values. See the FAQ Question about how to prepare your data to take advantage of this capability.

We can work with you to securely transfer and store your data for analysis.

After we have access to your data, we can import and run the analysis based on your desired clustering Goals.

Cluster Results can also be easily downloaded from a secure download link.

There are several results formats that you can download, they are as follows:

• Clustered Items (CSV File)
  
• Non-Clustered Items (CSV File)
  
• ALL Results (ZIP File)
  
  • finalMatches.csv - Clustered Items
    
  • finalMatches.json - Clustered Items in JSON Format
    
  • graphData.json - Plot Data of Cluster Profile (probalby not useful) in JSON Format
    
  • groupIndex.json - Items and Cluster Groups in JSON Format
    
  • noMatches.csv - Non-Clustered Items
    
  • --- VERY USEFUL FOR PRODUCT RECOMMENDATIONS --- (i.e. eCommerce)
    
  • groupIndex_Primary.csv - Primary Matched Components Index
    
  • groupIndex_Secondary.csv - Secondary Matched Components Index
    

The groupIndex files are VERY useful for directly loading your results into eCommerce sites or enterprise services that need to recommend similar or alternative items. The format can be easily used for direct or reverse lookup of items!

SimilarParts.ai DOES PROVIDE automatic matching and intersection of Lists or Arrays of Column Values within your data.

BUT, you have to help SimilarParts.ai recognize that the values in your Column contain a "List" or "Array of Values". Here is how you do it.

BEFORE you export your CSV File, you need to wrap each value in the column of List or Array values with "Hard-Brackets" (e.g. "[" and "]"). This will tell SimilarParts.ai that the values within your cell are a "List" or "Array" of Values.

List or Array Values only require to be separated by commas (",").

For example, THIS:

• Column : Material
  
• Value : 1M29390, 2B2993, 4M2998
  

Needs to become.. THIS:

• Column : Material
  
• Value : [1M29390, 2B2993, 4M2998]
  

--- OR ---

For example, THIS:

• Column : ProtectionLevel
  
• Value : SAE-1020, ASME-10022 +B, SAE-100200 +B
  

Needs to become.. THIS:

• Column : Material
  
• Value : [SAE-1020, ASME-10022 +B, SAE-100200 +B]
  

--- Handing Values with Commas ---
IF you are dealing with List or Array Data that already contains commas (","), then you need to quallify the values with Quotes.

For example, THIS:

• Column : Scent
  
• Value : "Pumpkin, Fresh Baked", "Pumpkin, Blazing", "Pumpkin, Sweet"
  

For example, THIS:

• Column : Scent
  
• Value : ["Pumpkin, Fresh Baked", "Pumpkin, Blazing", "Pumpkin, Sweet"]
  

Simply inform or send us of the new the CSV File and we will take care of the rest!

All of your Goals will use the updated file, you just need to re-run your Goals to obtain your new results!

SimilarParts.ai was designed with eCommerce in mind!

One of the biggest problem with eCommerce recommendations, they often NEVER show anything "similar" to what a customer is looking at, and "similar" is a highly debated subject in the industry.

Basically, we view "Similar" as this. --- ANYTHING YOU WANT IT TO MEAN!

So if you are looking to recommend things by:

• color
  
• size
  
• material
  
• function
  
• purpose
  
• unique attributes
  
• price, or
  
• whatever you think - even combinations of the above...
  

YOU CAN!

Since you only need a CSV File as your input, we also deliver CSV files as exports.

BUT - if your goal is to quickly take your results and import them for eCommerce recommendations, you should export ALL Results then find these files:

• groupIndex_Primary.csv
  
• groupIndex_Secondary.csv
  

These files are an "index" that you can quickly relate products that SHOULD be shown if a customer looks at any specific item that has either Primary or Secondary Matches within your SimilarParts.ai Cluster Results.

These files can be immediately loaded into your eCommerce platform - or a database if you build your own eCommerce and make use of the results immediately!

Don't worry - EVERYTHING within SimilarParts.ai is 100% Secure!

Everything -- your Uploads, Data Storage and Downloads are encrypted.

Delete anything it is GONE!

We do NOT backup your data. So be mindful of this security feature.

SimilarParts.ai is a Cloud-based Service.

There is NOTHING TO INSTALL, configure or license within your company or on your computer.

SimilarParts.ai clustering costs are computed based on the following:

• Number of Rows (Your CSV Data)
  
• Number of Columns (Your Goal Definition)

The above model enables you to pay for only what you need, based on the amount of your data and the complexity of the calculation based on your goal definitions.

YES!

If you or your company needs help analyzing your data, we can help!

The following services are offered:

• Data Clensing and Population
  
• Alternative Product Identification
  
• Data Clustering and Analytics
  
• Product Recommendation Building
  
• Identify Products for Consolidation
  
• Identify Products for Elimination
  

We offer excellent support and results for our customers!

YES!

YES!

If your clustering needs exceed the currently delivered capabilities, we can help. We can setup stand-alone, dedicated, clustering services for your organization.

If you need additional support email us at --> help@similarparts.ai