NU Seattle Student Makes Great Strides at Concur Internship

Yan Luo is just beginning his second year as a candidate for a Master of Science in Computer Science (MSCS) at Northeastern University’s Seattle campus. However, he’s already made an impact in his field through an internship at Concur, a travel and expense management company headquartered in Bellevue, Washington.

During the summer of 2017, Luo worked with members of Concur’s data science team to develop a program to identify the language of expense receipts that are sent to Concur’s system.

“One of the products that we provide our customers, ExpenseIt, allows them to create expense reports by simply taking pictures of their receipts instead of manually entering all required information,” explained Everaldo Aguiar, Luo’s mentor during his internship and also a part-time lecturer at Northeastern’s Seattle campus. “Those images are uploaded to Concur, where they are quickly processed and have important information extracted from them by a series of machine learning models. The user can then simply verify this information and submit their report for reimbursement.”

Through a process known as optical character recognition (OCR), the system can extract text data from these receipts images, and from that, important information such as the receipt’s vendor, location, and amount are extracted.

Recently, Concur has begun marketing their services outside of the U.S. Luo’s project allows the system to identify, based on the receipt image, what language they are in – making a particularly accurate distinction between Chinese, Japanese, and Korean. This, in turn, can majorly improve the quality of the OCR text

“For all the Japanese receipts we have so far, only a small percentage are recognized as Japanese,” Luo explained. “With my project, I can identify the language before receipts are processed and that allows us to get much higher quality text from them.”

Concur’s internship program is run in conjunction with SAP, a software and technology company. Luo spent most of the summer developing his project, which culminated in a presentation to Concur’s leadership in August. For this presentation, he received an SAP internship award, which highlights the top projects that interns developed over the summer.

Aguiar says that employees at Concur will continue to use the process Luo developed even though his internship has ended.

“Usually internship projects are designed to give interns exposure to practical skills, but their life can sometimes be restricted to that summer,” he said. “Yan’s project went beyond that and we will continue to mature the work that he started. This really is a testament to the quality of what he provided to our team .”

Luo’s success in his internship is an example of how invaluable summer interns can be to a team like Aguiar’s.

“Interns provide a fresh new set of eyes to look at problems we handle every day,” Aguiar said. “Students come in with a different perspective and they give us different ways of looking at these problems.”

In other words, experiential learning works both ways: students bring a fresh perspective to existing problems while also gaining valuable insight about the industry they hope to work in.

Luo, who got his undergraduate degree in biology in China, is excited that Northeastern offers him the opportunity to expand his real-world knowledge of computer science.

“I took my first machine learning course online but actually learned very little from it,” he said. “At Northeastern, machine learning was taught in a lot more detail.”

His skills have only improved through his internship at Concur and Aguiar’s mentorship.

“Everaldo told me about a lot of different techniques that I had never heard of before, like how to deploy my model to more servers,” Luo said of Aguiar. “Without him, I would do the project like a course with a report, but I would never make it work as good as it is right now.”

The experience at Concur has made Luo even more excited about his future career in data science.

“I feel very happy that I can use data science techniques in real life and make a difference,” Luo said.