Calendar management
Embedded contextual email
Scheduling automation
Automated outreach
Telehealth
Marketplaces
Recruiting
Document management
Real estate
Communications software
Docs
SDKs
Tutorials
API status
Changelog
Support
Blog
Events
Guides
Newsroom
Partners
Build vs. buy
Customer stories
Trust & security
Why NylasExtract Text from Images with the Nylas OCR Endpoint
Detect and extract text found within PDFs, TIFFs, and handwritten images with the Nylas Neural API.
Detect and extract text found within PDFs, TIFFs, and handwritten images with the Nylas Neural API.
While virtually every industry relies on crucial information found in digital images, recipients still spend far too much time and money extracting that information in order to make it trackable and actionable. Take billing invoices, for example – 28% of companies take 10-15 days to process a single invoice, and the bottom 25% performing companies spend $10 or more per invoice processed. Companies with complex processes can even spend up to $40 per invoice. When you process hundreds of invoices at a time, this process becomes incredibly laborious and expensive.
In addition to being costly, manually extracting data from images is error-prone in myriad ways – misplacing a crucial decimal or omitting or misspelling an entry causes the data to get lost entirely, or to present itself inaccurately.
Nylas built our OCR (Optical Character Recognition) endpoint to solve the challenge of manually extracting data from images. OCR is a feature of the Neural API that allows you to automatically extract text from images at scale, eliminating human error and adding valuable hours back into your users day.
Manually extracting information from images is both time-consuming and error-prone. Your users need a way to identify and capture data efficiently and correctly. Due to variability in different document types; however, building OCR capabilities in-house is complex and requires your teams to divert resources to build, train, and maintain the models.
As a feature of the Neural API, all you need is a single API request to convert images into parsed out, consumable information.
Emails synced through Nylas automatically assign a File ID to each attachment included in the email. Make a PUT request to the api.nylas.com/neural/ocr endpoint to intelligently detect and convert digital text information into machine-readable text.
Here, we’re asking the OCR endpoint to run OCR on an attachment:
curl -X POST 'https://api.nylas.com/neural/ocr \
-H 'Authorization: Bearer ACCESS_TOKEN' \
-d '{
"file_id":"sf4df7ra8"
}'
You can see the example response below where the OCR endpoints reads the invoice number and account number (abbreviated for readability):
{
"object": "file",
"content_type": "application/pdf",
"filename": "Invoice.pdf",
"id": "sf4df7ra8",
"ocr": [
"Summary of Charges\nThis document is used for expensing",
"Account #: 15830\nInvoice #: 84572",
"..."
]
}
Learn more about how the OCR endpoint works in our docs here.
OCR doesn’t just automate the process of invoicing and expense reporting. It has far-reaching applications across numerous verticals. We’ve included a few examples below:
Incorporating OCR capabilities into your application not only improves end user experience but also enhances its ML offerings. The OCR endpoint enables you to provide time and cost savings to your users without changing your roadmap.
Speak to a platform specialist now to learn more about OCR and other intelligence features of the Neural API like Signature Extraction, Clean Conversations, and Categorizer.
Please enter your email address and receive the latest updates.
Dominic is a Product Marketing Manager at Nylas. In his spare time, he loves to hike and go to the beach with his dog.
