Create API-fillable PDF templates in seconds with AI

At Anvil, we're committed (slightly obsessed) to making document automation as quick and painless as possible for companies with high volumes of paperwork processes. Our PDF Filling API helps product, engineering, and ops teams in many industries—HR, insurance, financial services, healthcare, and more—streamline their workflows by allowing them to create and fill PDF templates with just a few lines of code. The PDF filling API is easy to integrate once you have a PDF template set up, but setting those templates up often takes (lots of) time.

We've recently introduced new AI-powered features aimed at eliminating time and effort of setting up a PDF template, giving you an API fillable PDF template in seconds. This post is aimed at giving you a tour through the new AI features: Field Detection, Field Labeling, and AI-generated Human-Readable field aliases.

Here’s how they make your life easier:

1. Field Detection: smarter, faster, and more accurate

Imagine you work for an insurtech/insurance company. You deal with hundreds of different insurance forms, each unique depending on the US state. Templatizing (and maintaining) these PDFs steals weeks worth of your operations or developer colleague’s time.

Our new Field Detection feature uses cutting-edge computer vision to automatically identify all the fields in your document in seconds, whether it’s a scanned image, a Microsoft Word (docx) file, or a Google Doc. It recognizes underlines, form fields, and everything in between.

In the video below, you’ll see our Field Detection in action on a standard Non-Disclosure Agreement (NDA). With just a few clicks, all fields are automatically identified, demonstrating the simplicity of the process even on common documents you use every day.

2. Field Labeling: context-aware automation

Once the fields are detected, our AI goes a step further with Field Labeling. It scans the surrounding text to determine the best label for each field and links related fields together. For example, it can automatically recognize that two fields are meant for an employee’s first and last name or that a group of fields corresponds to an employee’s home address. This means your templates are not only filled accurately—they’re smart enough to handle complex, related data with ease.

Watch below as we take on a more complex document—a patient onboarding form. Our AI handles edge cases with dense fields and more sophisticated labeling, ensuring that related fields like names, addresses, and contact information are grouped and labeled intelligently.

3. AI-Generated Human-Readable Field Names: Easier API Integration

We’ve also enhanced how our API interacts with your templates by introducing human-readable field names (we call these `field aliases`). Our AI generates names for each field that are clear and easy to use. So instead of dealing with cryptic or confusing labels, you can interact with your PDF fields in a way that makes sense—leading to faster integrations and fewer errors.

Here's a before and after screenshot of how Field Names and Field Aliases get populated with and without our new AI features.

Save product and developer team's time with Document AI

These new AI features are all about saving your product and developer teams’ time and making your workflows more efficient. By automating what used to be manual processes, you can focus more on what really matters: building great products.

If you want to give Anvil a test drive (no signup required), upload any document file and watch the magic unfold.