Meet an Intuit Engineer – Justin Chiang

When the COVID-19 financial crisis hit, Justin and his team built upon the knowledge engine used for TurboTax to develop and launch Intuit Aid Assist on the Intuit Platform – a financial aid tool that has helped 660,000 small businesses and individuals navigate federal relief. How would you describe your role to another engineer? JC:

When the COVID-19 financial crisis hit, Justin and his team built upon the knowledge engine used for TurboTax to develop and launch Intuit Aid Assist on the Intuit Platform – a financial aid tool that has helped 660,000 small businesses and individuals navigate federal relief.

How would you describe your role to another engineer?

JC: I’m a software engineer working on the Futures team here at Intuit, where our goal is to create game-changing Technology, Design and Products to build a durable competitive advantage for Intuit and our customers. I work primarily as a backend engineer on the Knowledge Engine (KE) team, developing tools and infrastructure that make it easy for humans to codify their knowledge and creating algorithms that can auto-generate entire experiences out of those codified rules. A typical day of mine is usually spent collaborating with other engineers at architecture or coding sessions and working with teams across Intuit to improve and explore new use cases of KE.

What is the most rewarding part of your role? When do you feel most fulfilled?

JC: At Intuit everyone is “Customer Obsessed”, developing deep customer empathy through customer interviews and follow-me-homes. Our customer’s problems become simply our problems and we sweat every detail looking at possible solutions.

One notable example was an interview with a customer who showed just how difficult it was to figure out how much overtime to pay employees. The fiasco started with checking forums and websites to make sure you had the latest legislation updates, then reading through numerous labor bulletins and charts in order to be able to understand and apply the myriad of rules to determine whether an employee was even eligible for overtime (or fell under a specific exception, or an exception to the exception), before finally being able to apply the actual overtime pay calculation. It turns out that this process was so difficult that overtime disputes were a leading cause for labor lawsuits and only 40% of our customers were able to apply all the rules correctly. This actually led the team to improve Knowledge Engine and develop a series of tools which were used to codify very complicated overtime rules for all 50 states (plus Washington DC!) that our engine used to detect errors and notify customers, resulting in 95% of customers become compliant with the laws. It’s extremely fulfilling to not only see how the technology I’ve helped build has solved a customer’s problem but also realize how large of an impact reducing fear, uncertainty, and doubt has on our customer’s lives knowing that Intuit has their back!

What made you choose Intuit and why were you drawn to the company?

JC: What really set Intuit apart for me are the people and company culture. My first impression of Intuit was at an info session back in college that a friend had organized. During the talk, one presenter demoed how his team was using machine learning to take PDFs of tax law and automatically generate code that powers TurboTax. To make sure there weren’t mistakes that would result in a customer getting a tax bill instead of a refund, they leveraged data to generate millions of test simulations to run through the engine to ensure accurate calculations every time.  

I was amazed at how Intuit was using innovative tech to solve huge problems and as I was chatting with the presenter and other Intuit employees afterward I realized just how genuine and helpful they all were, candidly answering questions about Intuit or their roles that anyone brought up.  Having joined Intuit I can confirm there’s nothing closer to the truth; I’ve had tremendous support and coaching from fellow engineers, managers, and leaders throughout the company, all of whom are genuinely looking to help you succeed. Despite being a new grad, on my very first project, I flew with the team to interview customers and truly understand the problem before crafting and implementing a solution, one that has since helped millions of customers and also led to the filing of my first patent. Oh and the presenter at the info session? He’s now a fellow teammate and continues to be an amazing mentor and advocate.

On top of all that, being able to work in San Diego where it’s sunny 360 days of the year isn’t a bad perk either!

What were the most important immediate impacts of scaling this novel technology approach for Intuit and the Intuit platform? For Intuit customers?

JC: At Intuit, customers always come first. When the pandemic hit, more and more businesses were closing their doors each day and we knew we had to do something to help and fast. By leveraging Knowledge Engine, the team was able to create personalized, right-for-me experiences that helped each customer navigate the complicated set of federal aid rules, like determining if they’d be eligible for a loan or if a portion of it could potentially be forgiven. This was all done in record time with the first proof of concept built with KE in a matter of hours and the entire experience – including numerous legislation changes from lawmakers –  taking just 21 days. Internally, KE has been an integral part of Intuit’s AI-powered platform and scaling KE has increased artificial intelligence adoption across the company, allowing teams to move even faster and provide more personalized experiences to our customers.

KE technology has been around for some time, but with Intuit Aid Assist the team was able to innovate using it to rapidly meet new customer needs. As you look ahead, what about the Intuit platform do you think is the most promising?

JC: What’s really exciting is taking a capability first mindset and figuring out how to organize all the powerful services and experiences across the entire technology ecosystem into a unified One Intuit platform. By integrating everything together, we’re breaking down barriers across teams which really enables more collaboration and innovation while bringing speed to benefit. As a platform, innovations that a single engineer comes up with can benefit the entire company: Machine Learning models and KE powered experiences can be easily shared between teams, the ability to setup infrastructure to support new features can be done in minutes; the company is really coming together to give our customers the power of the entire Intuit platform.

What additional applications do you see for technology like KE or other tech on the Intuit platform?

JC: I see a lot of potential in more applications leveraging Intuit’s artificial intelligence, which we’ve defined as Machine Learning, Natural Language Processing, and Knowledge Engineering. Machine learning can be used to drive business insights or nudge customers to make smart financial decisions. Our learnings in natural language processing and understanding enable more natural chatbots that customers can interact with to explore their tax return, or even be used to turn a doctor’s new patient questionnaire directly into a KE experience. With Knowledge Engine, the team has already done a lot of amazing work in generalizing the engine to handle any domain regardless of whether it’s finance-related or not, and I can imagine KE being used to power immigration forms, or college and scholarship applications.  

How did the urgency around the impact of COVID-19 on customers accelerate your work on Intuit Aid Assist, Intuit’s financial aid tool for small businesses?

JC: I think this is best summarized by a Benjamin Franklin quote that “out of adversity comes opportunity”. The urgency around COVID really brought all of Intuit together to brainstorm ways to solve a pressing customer problem. After rounds of customer interviews and talking with government agencies, what we now call Aid Assist was originally estimated to take 8 months of effort, time which many customers simply did not have to wait for a relief package. This led the team to lean in on newer technologies like Knowledge Engine and its authoring tools and also prompted further improvements to KE by expanding the ability to auto-generate an interview experience from a set of rules. These innovations enabled a single tax analyst to transfer the knowledge from 880+ pages of the CARES Act into a spreadsheet, and leverage KE to automatically generate hundreds of screens and stitch them together into personalized interview experiences, reducing development time to just a couple weeks.

Despite constrained timelines, Intuit Aid Assist was built with the same level of rigor as any other Intuit product. Countless teams across the company contributed to Aid Assist and worked together to get Aid Assist into the hands of customers as quickly as possible.  Security teams checked for vulnerabilities while others ran performance tests to make sure the interview experience was snappy and smooth.  Content authors worked with analysts and legal teams to ensure accuracy while designers and engineers polished the experience.  Leaders worked to unblock teams and even Scott Cook, Intuit’s founder, could be found at Saturday review sessions to coach teams and improve the Aid Assist experience for our customers.

Looking back at you and your team’s success over the last year, what are you most proud of accomplishing? What are you most excited for in 2021?

JC: The team has definitely hit a bunch of milestones in the last year but if I had to choose one, it’s Intuit Aid Assist.  Helping 650K+ small businesses at a time they need it most is something to be proud of in itself – not to mention being able to do it in a matter of weeks, during a global pandemic with everyone trying to figure out how to work from home. I also have to give a huge shout out to the entire Intuit family; it wasn’t just the numerous employees across the company who volunteered time that made the project a reality but also their significant others, kids, pets, and loved ones who helped support them during these crazy times and enabled us to get Aid Assist out to our customers as quickly as possible.

Looking to the future, I’m excited to continue building upon the innovations and learnings we’ve gained working on KE and strive to make it even easier and faster for teams to leverage the entire Intuit platform and create brand new experiences from the ground up. Intuit is well on its way as an AI-powered platform and by making integration points even more seamless, it’ll enable teams to deliver more customer benefit even faster and truly power prosperity around the world!