Batch Data Matching API - File-Based Similarity Key Matching and Deduplication

See this API in action with our no-code, browser-based Matching Wizard

Try the full matching engine on your own files without writing any code. Upload a CSV or TSV, map your columns with a few clicks, and see your matches in seconds.

Launch the Matching Wizard

API Overview

The Batch Data Matching API processes an entire file in one request. It reads a CSV or TSV file, generates an AI-powered similarity key (SimKey) for the field or fields you choose in every record, and returns the records grouped into clusters that refer to the same real-world entity. It is built for high-volume deduplication and record linkage across customer, vendor, supplier, and account datasets, without writing per-record matching logic.

AI-Powered

Uses AI models and specialized algorithms to standardize and match company names, personal names, and addresses, including abbreviations, acronyms, and misspellings. It also works with international data and different character sets, matching names and addresses across languages, scripts, and regional formats.

Entire-File Processing

Point the API at a CSV or TSV file and it processes every record in the entire file in parallel, streaming back matched clusters. No need to loop through records or call the API once per row.

Flexible Match Functions

Match on company name, full name, or address individually, or combine company and address, company and full name, or address and full name for higher-precision linkage.

How It Works

Each record in your file is assigned a similarity key based on the field or fields you choose to match on. Records that resolve to the same real-world entity receive the same key, so grouping by key produces your match clusters. The example below shows company-name matching: inconsistent spellings and formats of the same organization all share one key.

Company Name (input)	Similarity Key
General Electric	`Xk7pR9vLmQ2wYdN1Ai4RfV0SRJf2dJwDO0C`
Gen Electric Corp	`Xk7pR9vLmQ2wYdN1Ai4RfV0SRJf2dJwDO0C`
GENERAL ELECTRIC COMPANY	`Xk7pR9vLmQ2wYdN1Ai4RfV0SRJf2dJwDO0C`

Microsoft Inc	`Hm3qT8bKzPfW5cxE5Y6ImDHMQUX6EMlBHD9`
microsot	`Hm3qT8bKzPfW5cxE5Y6ImDHMQUX6EMlBHD9`

Banco Santander	`Rn4sW7yPkLvJ9a0BfQx3mHcTdUeIgKoZlN`
Santander Group	`Rn4sW7yPkLvJ9a0BfQx3mHcTdUeIgKoZlN`

Records that generate the same similarity key are likely the same entity and form a match cluster. When you choose a combination function such as company-and-address, two records match only when both fields are similar, which reduces false positives compared with matching on a single field.

API Details

Endpoint

https://match.interzoid.com/match

Request Format

https://match.interzoid.com/match?apikey=[apikey]&connection=[file URL]&filetype=[csv|tsv]&function=[match function]&company_column=[n]

Parameters

Parameter	Required	Description
`apikey`	Yes	Your Interzoid API license key.
`connection`	Yes	Public URL (http/https) of the input CSV or TSV file to process.
`filetype`	Yes	`csv` or `tsv`. Sets both how the input is parsed and the output delimiter.
`function`	Yes	Match function: `company-name-only`, `fullname-only`, `address-only`, `company-and-address`, `company-and-fullname`, or `address-and-fullname`.
`company_column`	Conditional	1-based column position of the company name. Required for any function that matches on company.
`fullname_column`	Conditional	1-based column position of the full name. Required for any function that matches on full name.
`address_column`	Conditional	1-based column position of the address. Required for any function that matches on address.
`has_header`	No	`true` if the file's first row is a header to skip (not matched or counted). Default `false`.
`showkeys`	No	`true` appends the similarity key as the last output column. Default `true`.
`matchesonly`	No	`true` returns only clusters of two or more matching records; `false` returns every record with its key. Default `true`.
`includeheaders`	No	`true` writes a header row as the first output line. Default `true`.

Sample Request

https://match.interzoid.com/match?apikey=fh5hs7*****&connection=https://dl.interzoid.com/csv/companies.csv&filetype=csv&function=company-name-only&company_column=1&has_header=true

Sample Response (CSV, matches only, with keys and header)

company_name,simkey General Electric,Xk7pR9vLmQ2wYdN1Ai4RfV0SRJf2dJwDO0C Gen Electric Corp,Xk7pR9vLmQ2wYdN1Ai4RfV0SRJf2dJwDO0C Microsoft Inc,Hm3qT8bKzPfW5cxE5Y6ImDHMQUX6EMlBHD9 microsot,Hm3qT8bKzPfW5cxE5Y6ImDHMQUX6EMlBHD9

Clusters are separated by a blank line. Each cluster contains the records that share a similarity key.

Example Calls

The same endpoint supports several match functions. Below are common patterns, each pointing at a sample file. Replace the file URL and column positions with your own.

1. Company Name Matching (CSV)

Deduplicate a customer or vendor list by company name. The company name is in column 1, and the file has a header row.

https://match.interzoid.com/match?apikey=fh5hs7*****&connection=https://dl.interzoid.com/csv/companies.csv&filetype=csv&function=company-name-only&company_column=1&has_header=true

2. Full Name Matching (CSV)

Find duplicate individuals in a contact list by full name, where the name is in column 1.

https://match.interzoid.com/match?apikey=fh5hs7*****&connection=https://dl.interzoid.com/csv/peoplenames.csv&filetype=csv&function=fullname-only&fullname_column=1&has_header=true

3. Address Matching, All Records Returned (CSV)

Match a file by street address in column 1, returning every record with its key rather than only matched clusters. Set matchesonly=false to see the key assigned to each row.

https://match.interzoid.com/match?apikey=fh5hs7*****&connection=https://dl.interzoid.com/csv/addresses.csv&filetype=csv&function=address-only&address_column=1&has_header=true&matchesonly=false

4. cURL Example (API key in header)

The API key can be passed in the x-api-key header instead of the query string. The matched output streams to your terminal or a file.

curl --header "x-api-key: fh5hs7*****" "https://match.interzoid.com/match?connection=https://dl.interzoid.com/csv/companies.csv&filetype=csv&function=company-name-only&company_column=1&has_header=true" -o matches.csv

Individual Matching API Product Pages

Each match function used by this batch API is also available as a single-record matching API, with its own documentation, examples, and interactive tools. Explore the individual product page for any function:

Company Name Matching

Match records by company or organization name (company-name-only).

View API

Individual Name Matching

Match records by personal or full name (fullname-only).

View API

Street Address Matching

Match records by street address (address-only).

View API

Company and Address Matching

Match records by both company name and address (company-and-address).

View API

Company and Full Name Matching

Match records by both company name and full name (company-and-fullname).

View API

Address and Full Name Matching

Match records by both address and full name (address-and-fullname).

View API

Try the Request Builder

Build a Batch Match Request

Fill in your file URL, match function, and column positions to generate a ready-to-call request URL. The batch API streams a file of results, so this tool builds the URL for you to call with curl or the Match Wizard rather than running it inline.

API Key

File URL

File Type

Match Function

Company Column

Full Name Column

Address Column

Algorithm & AI Model Information

Matching Algorithms

The company name component uses the model-v4-wide algorithm for AI-powered organization name standardization with wide matching, including abbreviations, acronyms, and international company names. The address component uses the model-v3-wide algorithm for AI-powered address standardization with wide matching (street number and street name, ignoring unit numbers for broader matching). Full name matching uses AI-powered personal name standardization. Combination functions require both component fields to be similar before two records match.

Note: Because the API returns a similarity key per record, you can match within a single file or, by combining files, match across datasets. Records that share a key are candidates for merging, deduplication, or linkage.

Code Examples, Database Queries, and Specifications

Usage Examples

See quick and easy code examples in multiple programming languages, including Python, Node.js, Java, Typescript, and Go. Also see how to use the matching APIs within database queries on popular database platforms and open specifications such as OpenAPI.

See All Examples

Additional Resources

Use this API with the No-Code Match Wizard

Our Match Wizard no-code batch tool runs this matching engine within our parallel-processing Cloud platform. Upload or connect a CSV/TSV file, map your columns with a few clicks, and match within or across datasets in seconds, without writing any code. It is the fastest way to use this API with higher volumes of data.

Start Now

Match Directly Against Snowflake and Azure SQL Tables

In addition to files, the matching engine can read directly from Snowflake and Azure SQL tables, generating similarity keys for table data using the same match functions. Try it with our no-code Snowflake Matching Wizard, which connects to your warehouse, maps your columns, and matches table data in seconds.

Launch the Snowflake Matching Wizard

Alternative Access: Pay Per Use with x402 (No API Key)

Most users access this API with a standard Interzoid API key, as shown throughout this page. As an alternative, the same endpoint also supports x402 pay-per-use payments, where you pay per record in USDC on the Base network instead of using a key or subscription. This is aimed at AI agents and automated systems that need on-demand access without managing credentials. If you are using an API key, no action is needed.

Learn About x402 Access

Service Bureau - Our Data Matching Task Force

Want us to do it for you? We can take your data and run our matching products on your behalf with a quick turnaround:

Contact Support

Video Description

Watch a short video describing similarity keys and how they work.

YouTube

Sample Matches Found Using Similarity Keys

Company	Similarity Key
Toyota Corporatin	`tmOnl2ryWlUguo7nebVhQ4WQnQ4PvxAd81MHwa_y_to`
Tyota Corp (Japan)	`tmOnl2ryWlUguo7nebVhQ4WQnQ4PvxAd81MHwa_y_to`
Toyota Corporation	`tmOnl2ryWlUguo7nebVhQ4WQnQ4PvxAd81MHwa_y_to`

BMW	`qLAoXfmrTOrT8lcGlrZ2bDELo2swEzXmg53zLWdOyqQ`
Bayerische Motoren Werke AG	`qLAoXfmrTOrT8lcGlrZ2bDELo2swEzXmg53zLWdOyqQ`

Banco Santander	`i5I9QIUydH_dP6ldIuZjJrhGpp80PJNSbbJNxwj5APc`
Santander Group	`i5I9QIUydH_dP6ldIuZjJrhGpp80PJNSbbJNxwj5APc`
Banco Santander S.A.	`i5I9QIUydH_dP6ldIuZjJrhGpp80PJNSbbJNxwj5APc`

Microsoft Inc	`xUhcrilUNsRiCthe7rXkIupHiCbhhgyLrKNAcXruwoA`
microsot	`xUhcrilUNsRiCthe7rXkIupHiCbhhgyLrKNAcXruwoA`
MICROS0FT	`xUhcrilUNsRiCthe7rXkIupHiCbhhgyLrKNAcXruwoA`

IBM	`edplDLsBWcH9Sa7ZECaJx8KiEl5lvMWAa6ackCA4azs`
Intl businessmachines	`edplDLsBWcH9Sa7ZECaJx8KiEl5lvMWAa6ackCA4azs`

Amazon.com	`tyGzXZjfZUqhgqt6mqNZF8MCsn-QQV1NJbysxSTB7aI`
Amazon Incorp.	`tyGzXZjfZUqhgqt6mqNZF8MCsn-QQV1NJbysxSTB7aI`

PetroVietnam	`gVzS9hXt8FdAISVdnB7uisfT2DzIzRfGKgdKz2JmnJc`
Vietnam Oil and Gas Group	`gVzS9hXt8FdAISVdnB7uisfT2DzIzRfGKgdKz2JmnJc`

Groups of rows with the same color share the same similarity key, indicating they are considered the same entity. The same approach applies to full names and addresses, and to combinations of fields.