How H&R Block uses Azure AI to transform tax returns
H&R Block customers know they can trust our tax pros to provide help, guidance and expertise. And we know technology is key to providing a seamless customer experience, whether in-person with tax pros or through our DIY product. That’s why we’re happy to have partnered with Microsoft to add new technology just in time for tax season. This case study by Microsoft demonstrates how we, together, have doubled down on our commitment to delivering a more personalized, streamlined tax preparation experience.
A new job. A change in relationship status. A move across the country. H&R Block understands that tax returns are personal—a diary of your year’s most impactful events. When the organization’s 21 million clients rely on the tax professionals at H&R Block, providing individualized support for each client’s unique needs is paramount. And whether clients come in with a shoebox full of tax forms or file online, H&R Block is using Microsoft technology to transform the tax filing experience. With AI-powered services like Azure Form Recognizer and Azure Cognitive Search, H&R Block tax professionals can spend more time building meaningful, personalized client experiences—and helping each client get the most out of their tax return.
Transforming tax time for a digital age
In 2021, the 60,000 tax professionals at H&R Block prepared more than 21.6 million tax returns. That’s 21.6 million individuals with varying needs, concerns, complex circumstances—and each with a shoebox full of tax forms.
At least, “shoeboxes full of tax forms” used to be the dominant imagery of tax season. But as the world becomes increasingly more digital, the needs and preferences of clients are evolving fast. “For many people, taxes are a time of anxiety,” says Aditya Thadani, Vice President at H&R Block. “We understand that it’s not just about tax preparation: it’s about financial health and feeling confident about getting the expertise and help they need.”
While some clients still prefer to meet in-person and work with paper documents, others want to prepare their taxes online, or take a hybrid online and in-person approach. No matter how clients choose to connect, the tax professionals at H&R Block know tax time isn’t everyone’s favorite season, so they want to focus their time on getting to know each client, learn about their needs, and find the best solutions for them.
That said, anyone who has ever done taxes knows it can require entering and making sense of a lot of data—and data entry can be a time-consuming and tedious effort, further complicated by the intricacies of the United States tax system. To help make the experience easier, less stressful, and more personal for both clients and H&R Block tax professionals, the company wanted to minimize time spent entering client information by equipping its team with a full history of each client, fast.
Shifting from stressful to streamlined
Today, most clients engage with H&R Block just once a year and roughly one-quarter of the organization’s business is conducted in the span of just a few days—so, every moment of client interaction matters. By removing silos in data and going fully cloud native with services such as Azure Cosmos DB and Azure Data Lake, H&R Block was able to unlock previously untapped AI capabilities.
With the power of Microsoft Azure Cognitive Search, each tax professional can find detailed client history in seconds. It’s a fully managed, omnichannel search experience that connects information from previously siloed sources—even if a client is moving from online to in-person consultations, or previously met with H&R Block pros in a different city.
Retrieving thorough client records from a centralized source helps tax pros understand the story fast. “If we don’t know who you are, we can’t really serve you well,” says Thadani. “Accessing useful information quickly is critically important to providing a differentiated client experience, which is why we use Azure Cognitive Search.”
Comprehensive access to a client’s past information allows H&R Block to focus more on the present. “Having information ahead of time allows our tax professionals to be better prepared to give clients a personalized experience,” says Erika Algarin, Director of Assisted Tax Transformation at H&R Block. “Our clients are often talking about their most private information—finances, bank statements, how much money they make. Having a tax pro that’s familiar with your circumstance, who understands what has happened in previous years’ tax returns, immediately helps them feel at ease.”
Making sense of the shoebox
Understanding a client’s journey comes down to having the right data—and a single tax return can have up to 100,000 fields of information and data points. To add further complexity, tax information comes to H&R Block in various forms—digital, the original document, a picture of a form—and tax forms change every year, sometimes right before tax season.
While ensuring accuracy of data input is essential, H&R Block tax professionals want to spend time connecting with clients, not typing data into form fields. The organization found a way to streamline the tax form data entry process with Microsoft Azure Form Recognizer, an AI service that uses machine learning to extract text, numbers, and critical information from a wide variety of tax documents, starting with W-2 forms, 1099s, and more.
“Give us a shoebox of tax documents, and we’ll use AI and machine learning to put the data in the right places,” says Sameer Agarwal, IT Director at H&R Block. The powerful machine learning classification model understands the language and patterns of tax documents. Then, Form Recognizer organizes stacks of papers into useful, digital information, and the Azure environment ensures sensitive data is handled in a highly secure manner.
Training these AI models to learn and adapt to the intricacies of tax documents enables H&R Block to operate with more accuracy and agility. The digitalization efforts also align with the company’s goal to explore more sustainable, environmentally friendly solutions by reducing paper usage.
And it’s a scalable solution: In future tax seasons, H&R Block will use this cloud-native Azure solution to classify more than 30 million tax documents per year. And with huge fluctuations in work volume throughout the year, trust in its technology solutions is critical—especially during the busy make-or-break days of tax season. “We conduct about 25 percent of our annual business in a matter of days,” says Thadani, “so having the ability to ensure operational reliability of our systems is core to our success.”
Doing more with less
With AI-powered processes in place, H&R tax professionals are finding data entry and retrieval more efficient. They can spend their time in more valuable ways: delivering differentiated client experiences and helping clients approach tax season with confidence. “Machine learning frees up the precious time we have with the client to actually have richer conversations about their financial needs and aspirations, beyond the immediate task of ensuring the best tax outcome for the current year,” says Thadani.
The solutions are built on the existing Azure environment within H&R Block, and Agarwal appreciated the way Microsoft sought to understand the complexity of tax forms. “There’s a lot of intelligence needed to solve this problem, which is why we chose Microsoft as our technology provider,” says Agarwal, “and the technology just works.”
“It’s our job to make sure that we’re proactively evolving our approach to taxes,” says Algarin. “Technology allows us to do that and helps us put our clients in the driver’s seat.”
Millions of American lives have changed during the pandemic—and H&R Block has a front row seat to it all. See how people just like you are learning to thrive in these changing times.
H&R Block releases new insights report outlining how Americans have persisted through the COVID-19 pandemic.
H&R Block announces new vice presidents of Government Relations and Technology. Meet the new members of the team.
H&R Block has filed a lawsuit against Block, Inc., formerly known as Square, Inc., seeking to enjoin the latter’s infringement of Block’s family of trademarks.