Trove
Headline · Coding · Leaning · Business
Trove is a free transaction enrichment API from Headline. It turns messy bank transaction data into clean merchant data, spending categories, and location details.
About Trove
What Is Trove
Trove is an API that reads bank and credit card transaction descriptions and returns the merchant behind them. Send it a string like "UBER TRIP 8005928996 CA", and it replies with the company name, domain, founding date, industry, business categories, headcount range, headquarters location, and official social profiles.
The problem it addresses is familiar to anyone building finance tools. Card statements arrive with cryptic strings full of reference numbers and location codes, and none of that tells a user where their money went. Trove replaces that noise with structured merchant data that an app can display, group, or analyze. Teams often reach for a merchant data API to solve this, and Headline trained a language model directly on financial transaction text rather than reusing a general purpose one.
The main limit is scope. The service is tuned for US transactions written in English, and the company states that other languages and regions may match less accurately. Bulk requests also cap how much work you can queue at once.
Getting Started
- Create an account on the Trove site and generate a personal API key.
- Send your key in the
X-API-KEYheader on every request. - POST a transaction to
https://trove.headline.com/api/v1/transactions/enrichwith thedescription,amount,date, anduser_idfields. - Read the JSON response for the matched merchant, or submit batches through the bulk endpoint and poll for results.
Product Information
A quick look at Trove's pricing, supported platforms, and performance.
Best for
The users, tasks, and scenarios where this tool fits best.
Users
- Personal finance app developers
- Fintech teams building spend dashboards
- Data engineers cleaning transaction logs
Tasks
- Cleaning merchant names from card statements
- Grouping spending by category
- Filling gaps in transaction records
Scenarios
- Building a budgeting feature
- Investigating where a payment went
- Correcting a wrong match
Key features
Real-Time Merchant Enrichment
The core endpoint takes a transaction description, amount, date, and user identifier, and returns the merchant plus its profile in a single call. Headline trained a language model on financial text specifically for this parsing job, so it handles the asterisks and reference codes that break simpler matching rules.
Rich Merchant Profile Fields
Each successful response carries a company profile: name, domain, type such as Public Company or Privately Held, industry, business categories, approximate size, and headquarters city, state, and country. It also returns official Facebook, X, and LinkedIn URLs, which saves you a second lookup step.
Bulk Enrichment Endpoint
The bulk route accepts an array of transactions and returns a request ID you poll for results. You can keep three batch requests in flight at once and submit more as they finish, which fits nightly jobs that process large card datasets in chunks.
Feedback Loop
A dedicated feedback endpoint lets you send back a corrected domain for any description you enriched. That closes the loop on wrong matches and, over time, improves results for the exact strings your app sees most.
Language Model Built for Transaction Text
Rather than routing financial strings through a general model, Headline tuned its own model on transaction data. The practical effect is better handling of abbreviated merchant names, embedded store numbers, and location suffixes common to card processors.
Pros and cons
Pros
- No cost to start: the service is offered free, with an API key created in a few clicks.
- Returns a full merchant profile, not just a cleaned name, including domain, categories, and headquarters.
- Bulk endpoint handles large volumes and reports progress through a request ID.
- Feedback endpoint gives you a way to correct and improve matches for your own data.
Cons
- Tuned for US transactions in English, so accuracy drops for other languages and regions.
- Bulk requests are limited to three concurrent submissions, which slows very large jobs.
- Requires an account and API key, so there is no anonymous trial of the endpoint.
Frequently asked questions
Trove takes a raw bank or credit card transaction description and returns the merchant behind it. The response includes the company name, domain, industry, categories, size range, headquarters, and social profiles.
Related content
Explore related tools, skills, and articles for Trove.
Trove Alternatives
Base Code
Base44 (Wix) · Coding · ProductivityBase Code is a shared cloud development environment from Base44. Teammates describe changes to an existing GitHub codebase in chat, preview them live and send them back as pull requests.
Google EmbeddingGemma 2
Google DeepMind · Writing · Image · CodingGoogle EmbeddingGemma 2 is a 740M-parameter open model that turns text, code, images, audio, and video frames into vectors in one 768-dimension space, running on phones and laptops.
GitGlow
GitGlow · Coding · ProductivityGitGlow is a local-first desktop Git client that lets independent developers review their own code and their AI agents' changes in one place before shipping.