Revolutionizing Document Management: How DocparserAI Transforms Data Extraction and Usability

Like many businesses today, at SureSwift we're using AI more and more, both in our product offerings and productivity tools for our teams. Some of our portfolio businesses are a natural fit for using AI to increase the value of their products to customers. A product powered by AI in thoughtful, smart ways can help save users time and effort with smoother functionality, intuitive features, and an overall better experience.

Docparser is one portfolio company doing this really well. We sat down with Lindsay Thompson, General Manager, and Kevin Strasser, Head of Product, to learn all about what they've been up to and how AI is improving Docparser for its users.

Keep reading to learn about Docparser, its case for using AI, and the neat AI features the team recently released along with their upcoming plans.

Table of Contents

1. What is Docparser?
2. The case for AI
3. DocparserAI: Docparser's AI engine
4. Changing the game with a much-improved user experience
5. Future plans

What is Docparser?

Docparser is a document processing and data extraction tool designed to help businesses and individuals extract data from documents. It extracts and transforms information from several different document types (like PDF invoices, purchase orders, bills of lading, bank statements, work orders, and more) and turns it into structured data (in database structures like .xls, Google Sheets, etc.)

The product is all about helping users optimize their processes to save time and effort — no matter what role or industry they're in. Every document has its own requirements and every company has a different use case, and Docparser accommodates it all.

Docparser allows users to:

1. upload any kind of document, like a PDF, an Excel spreadsheet, or a CSV,
2. quickly identify the information they want to take from the document (like names, addresses, invoice numbers, tables, handwriting, checkmarks, and more),
3. easily adjust, edit, and filter AI outputs, and
4. export that information where it needs to go (like in Excel, Google Sheets, CSV, or JSON format, or integrated with other platforms via API or webhooks).

This simple process is very malleable and customizable in lots of ways. But, the team knew they could do even more for customers with the release of AI technologies.

The case for AI

Thompson points out that with the introduction of AI technology to the market, customers are looking for optimizations within all of their interactions with SaaS products. In our industry, this looks like improvements in usability, efficiency, and accuracy of data extraction, to help further optimize business processes.

She continues, "With the introduction of AI technology, our customers can accomplish their process improvement goals more efficiently than ever. Our SmartAI Parser scans uploaded documents and automatically creates parsing rules to extract the exact data needed. Customers review the data and can accept, remove, or further customize the extraction to get the precise information they need, in the format that works best for their use case."

Docparser is looking at multiple different ways to improve the customer experience by adopting various AI technologies. By remaining technically flexible and curious in the AI space, the team can be innovative across more of its existing product infrastructure. The ability to customize the product to a specific user's needs has always been a huge value proposition for Docparser customers. They want to continue to offer this customization while leveraging AI to improve usability, efficiency, and product offering expansion. 

One way the Docparser team is getting creative with AI is expanding into unstructured data extraction. Their upcoming ResumeAI Parser release, for example, will allow HR professionals, recruiters, and hiring managers to extract information from hundreds of resumes regardless of their individual formatting. With extracted resume information available in a structured table, evaluating candidates, identifying top recruitment potentials, and/or organizing applicants will be fast and easy. 

DocparserAI: Docparser's AI engine

This month, Docparser is releasing DocparserAI, their combination of AI technologies within their flagship product, Docparser. Think of it not as one specific thing, but rather as the engine powering Docparser, helping to propel the product into the modern era of product development.

Some new functions have been developed using DocparserAI, which will be released soon.

1. SmartAI Parser (template)

The team's first DocparserAI release is a template that automatically identifies data to extract from a document. DocparserAI scans your document, identifies the type of data in the file (tables, charts, handwriting, etc.), and then creates rules for users to accept, modify, or delete, depending on what they need. 

‍

‍

This is a big change from the previous experience when customers would have to identify each item they wanted to extract and select the rule to process it correctly. While they had a wide variety of rules to choose from, setting up a parser for longer documents took time.

All users, regardless of experience, will get the SmartAI Parser template included in their account. This includes everyone from individual customers using Docparser for a few documents to large enterprise accounts leveraging Docparser to manage key business processes.

Flexible, customizable rules and new functionality: 'Like a Swiss Army knife'

The great thing is that even after SmartAI Parser has created rules and done a first pass on a document, it's completely customizable to the user's needs. They can apply filters, edit, rename, change the order, or even delete rules if, say, those rules aren't quite giving the output needed. There's tons of flexibility to edit, change, and do many things. "I think that part's really powerful compared to some of our competitors," Strasser notes.

SmartAI Parser also offers new functionality by handling handwriting recognition and receipt parsing, which were never supported before. For years, the product couldn't handle these things, and the team would have to send customers away.

Thompson points out a neat handwriting recognition example she recently saw in action: "A customer had a handwritten bus schedule. Everybody had written in a line — for example, on Monday this is the bus time, the destination, etc. The AI picked up everything perfectly."

SmartAI Parser also supports automated table recognition. "We're looking to add a separate rule so people can get this functionality in their other templates without having to actually use the SmartAI template," Strasser explains. "At the moment, you have to manually draw the columns through the tables to determine column headers, whereas it's all automated in the SmartAI template."

When describing SmartAI Parser, Strasser says, "It's like a Swiss Army knife. It handles any kind of document and does at least a 99% solid job where you might have to tweak just a few little things if it's really complicated."

Easier and faster than others

Something that sets SmartAI Parser apart from the competition is that you only need to upload one document. Other companies use a machine learning model that may require around 100 versions of a document to use properly. But with just one document, SmartAI Parser is much easier and faster.

2. ResumeAI Parser (template)

When it comes to resumes, the trouble is they're all formatted differently, with various data, styling, number of pages, and more. There's so much variety, which has always been a challenge for human capital teams wanting to easily get that output. It requires more time to sort through and review each document. To help alleviate this, ResumeAI Parser was created.

ResumeAI Parser allows users to upload resumes in any shape or form, and then extract the necessary information into a consistent structure. Users feed the tool any resume and it will extract necessary data (like name, email, social media accounts, experience, awards, educational background, skills, etc.) For example, recruiters can upload hundreds of different resumes, extract only the information they want, and then qualify candidates, creating rules to say, "I only want to look at candidates with X skill set." Then, this qualified candidate data is easily outputted into a file like a CSV or sent to a backend system. 

"Down the road, we'll probably work more with human capital teams to understand what else they might need," says Strasser. "There's a lot of potential to do interesting things, like scoring metrics when seeking candidates highly skilled in a certain area."

3. Filter for content summarization

Docparser will soon release a new content summarization filter, which consumes large amounts of data and produces a summary. It's quite flexible with several preset settings, yet everything is also fully customizable.

"The initial use case for this was a legal customer explaining they have these very large, long legal documents that are hard to sift through, and they just needed a summary," Strasser explains. "But, because it's legal, there's certain terminology that needs to be classified in a particular way."

Marketing is another area that will benefit from this tool. Marketers can take large amounts of text and get, for example, a paragraph summarized for a social media post or press release.

Strasser tells the story of a colleague sharing with a customer a very complicated table that included many columns, handwritten notes, and checkboxes. "He raved about how he couldn't believe it actually parsed the whole thing and outputted all of the data. (The tool) really is powerful and can save people a ton of time."

4. Smart checkboxes (rule)

Some users work with lengthy forms day in and day out. In the past with Docparser, if you uploaded a form with 150 checkboxes, you had to get 150 manual rules for the output of each.

The great thing is that now, you can use smart checkboxes once and the tool will automatically find all 150, create a rule for each one, and determine their output (e.g. whether a box is checked or not). This is much faster, especially for lengthy forms.

Changing the game with a much-improved user experience

The neat thing about Docparser's new AI features is that before, customers had to create their own rules and tell the product what to do. They had to know a bit about how to use the product and the process was more manual.

When asked about how DocparserAI will help improve that user experience, Strasser says it'll help new customers who get intimidated when they're first introduced to the product, which can often happen with so much functionality and customizability in front of them.

"We're hoping this helps with that. Since they can easily upload a document and see a result quickly, I think that's pretty powerful," he adds.

As well, Strasser says that many of us are busy and some people can be fickle when they've never used a certain product — they don't understand things right away or they're not sure if it really works for their purposes. "It's human nature when something isn't perfectly intuitive for how our brains work to easily get frustrated and give up," he points out.

"We've seen that a lot historically with trial customers that don't convert, but it's not like they've had a bad experience: When the data shows they only uploaded one document, we know they didn't really use the product — so, of course, they didn't see its value."

Future plans

The team will be releasing a resume parser in Docparser’s sister product, Mailparser, soon and is looking to add more AI features to that product as well.

As for what's next for Docparser, they're adopting AI every day and identifying processes and features to improve — whether it's parsing data from voice files, data validation, or anything else. Stay tuned!

‍

Learn more about Docparser and Mailparser, or sign up for SmartAI Parser and ResumeAI Parser to see how they can work for you and save you time and effort.

‍