When Akshay Singh first started filing his taxes after university, everything was simple.
But after marriage and kids, life got more complex and his tax equations grew more complex and daunting too. The once straightforward paperwork evolved into what felt like an overwhelming challenge, filled with new deductions and credits that were difficult to navigate.
“Over time, I totally lost the confidence to file my own taxes,” Singh said.
“It’s a challenge, a huge opportunity, and a great learning journey.”
Akshay Singh, Intuit
Today, Singh is an Engineering Group Manager at Intuit.
Intuit’s AI-driven expert platform powers TurboTax, Credit Karma, QuickBooks, and Mailchimp, which are used by approximately 100 million people and businesses around the world.
From that vantage point, Singh has seen firsthand how many people struggle with the same uncertainty, and now he has the opportunity to improve the technology that helps people file taxes, manage payroll, track business expenses, and more.
For decades, Intuit’s do-it-yourself financial technology made tax filing easier for those with simple returns. But for people like Singh with more complex situations that could benefit from additional guidance from a financial expert, the company knew it had to find an effortless way to connect people to the expertise they needed, and help them feel more confident in their finances.
Enter Intuit’s Virtual Expert Platform (VEP): a capability that combines AI, expert advice, and a decades-long understanding of customer needs.
With VEP, Intuit doesn’t have to wait for customers to ask for help while using their platform. Instead, it is able to anticipate their needs and proactively offer solutions, through a unique hybrid of AI and human expertise. The platform uses AI to connect people with the right tax or bookkeeping expertise when they need help, making expert support faster and more accessible.
Intuit is further enhancing its platform through an increased focus on delivering “done-for-you” agentic AI experiences, where the company automates tasks, end-to-end workflows, and entire functions, and connects customers to one of its more than 12,000 AI-enabled human experts for that last mile or to complete all of the work.
Intuit automatically uploads data the customer has connected, saving experts from manual data entry and searching various forums for customer details to make the first interaction a standout experience.
Intuit’s experts can complete the customer’s tax return in as little as two hours, offering proactive and personalized assistance, and providing opportunities for customers to access their money instantly.
“Taxes and finances are complicated, and our customers want expert help without the hassle,” said Shilpa Reddy, who leads the VEP as Vice President at Intuit. “Our platform helps customers get the right level of expertise, whether from AI or a human expert, so they can maximize their money with less effort and greater confidence.”
A chance for true ownership
The Intuit engineers who work on VEP were drawn to this work for different reasons.
Singh was drawn in by the appeal of the problem itself, and the chance to build technology that is used by millions of people each year.
Nicholas Yee, on the other hand, was drawn by the potential of building something foundational for a company like Intuit, in the heart of Toronto’s tech scene. Yee, who had previously worked in banking, startups, and scale-ups, saw VEP as an opportunity to work with AI and shape a team and product from the ground up. Today, he is a Senior Software Engineering Manager with the VEP team.
“I was excited to be part of a growth organization and get to help build out a team, build out a culture, and do that in Toronto,” Yee said.
When he joined, his focus was on finding people eager to push AI forward and experiment quickly. He wanted to shape the product, test ideas with real users, and make decisions that directly impact how millions of people access financial help. At Intuit, that’s exactly what he found.
“One of the things about my team is that we’re told to be experimental, to try things out,” Yee said. “I thought that was a great opportunity, since it allowed us to work at the speed of a startup, with the scale of a globally trusted brand.”
Experiments at Intuit are driven by data. Behind the scenes, teams analyze real-time customer interactions to identify the next problem to tackle and where AI can make the biggest impact.
The system behind VEP evaluates availability, expertise, and urgency before routing a case. If AI can resolve the issue, it steps in. If not, VEP escalates the request, connecting the customer to a tax or bookkeeping professional instantly. AI also enhances the expert experience by providing relevant insights and support to improve interactions with customers.
AI plays a role in the process, but it doesn’t replace the human connection that expertise depends on. “We are building so that humans will always be in the loop in the way we think about AI,” Reddy added.
This interplay between machine learning and human support is the heart of the VEP model. Intuit wanted to build a system that could predict customer intent, and surface the right kind of support, whether that is AI-driven assistance or access to real human expertise.
“We’ve been able to build AI-powered features to help our experts focus on the conversation with the customer by automating tasks like note-taking, searching for relevant resources that they can share with the customer, or even making recommendations on the best way to solve the customer’s issue. Our goal is to ensure that our experts have all the tools they need to provide the best customer experience,” Yee added.
For engineers, working on something like Intuit’s Virtual Expert Platform is a rare challenge. The stakes are high, the scale is massive, and the impact is immediate.
A culture of experimentation
Experimentation has shaped Intuit’s approach to product development from the start. In 1983, co-founder Scott Cook launched “Follow-Me-Homes,” a practice where engineers observe real customers using the software in their own environments. The idea was simple: watch, listen, and learn, rather than assume.
“We as engineers actually get to see experts using the product, which is a very unique opportunity,” Yee added. “These are also really insightful. Hearing what they like, what they don’t like, and seeing what their experience is like—it really helps me get a better understanding of the customer.”
Follow-Me-Homes continue decades later alongside Intuit’s Design for Delight framework, ensuring teams replace guesswork with firsthand insight. Intuit employees—not just engineers or designers—have access to training on design thinking as part of the company’s focus on customer-driven innovation.
Engineers are also embedded with product teams and designers to make decisions that define how financial expertise is delivered. This requires strategic thinking about AI, including how the platform analyzes customer queries to understand their intent.
Singh sees this evolution in AI’s abilities as foundational to the bigger opportunities ahead.
“We are really excited and deeply invested in making more AI-native experiences,” he said. “We’ve been investing in AI for more than a decade, and even with the experiences that are serving users today, we are getting an opportunity to rethink them in a new way.”
The team is still experimenting, refining, and figuring out how to push the platform forward. For Singh, that’s what makes the work exciting.
“It’s a challenge, a huge opportunity, and a great learning journey,” he said.
PRESENTED BY
Intuit is building the future of AI-driven expertise and is looking for engineers to help shape it. Explore opportunities on Intuit’s career site.
All photos provided by Intuit.