This week Innotech Capitals, an international private equity and venture capital fund, is expected to announce it has made a seed investment in Glever.com, a company that uses AI technology to make resume writing simple and effective. Nothing is more important – or painful – as writing a compelling resume. Fortunately, after some input from you, Glever automates the process and helps create a finished resume.
But that is just the tip of the iceberg, as founder, Xiaoxin Aaron Yin (pictured left) explains.
Q: What is the thinking behind Glever?
A: It is not easy to create a resume, and elaborating on work experience with emphases on just the right skills can be a daunting task for many. Glever uses AI technology to make resume writing as simple as several mouse clicks. It first asks the user to provide his/her job title, and then starts creating possible descriptions for the job. When receiving more hints such as skills used in this job, or one sentence describing the main responsibility, Glever can create more detailed copy, specific for what the user needs. It can even finish a half-written sentence. So, a professionally written resume with the right content can be written in minutes, instead of hours or days. Did I mention it is free?
Q: Tell me more about the underlying technology?
A: It is based on latest deep learning technology. We retrieve millions of professionally written resumes from the web, and build a deep neural network from them. This neural network learns how people write resumes: the wording, the skills, and the flow of text. When writing a resume, Glever looks at the information provided by user, such as job title and skills, and produces contents just as a live expert would.
Q: Can that tech be applied to other applications?
A: Automated writing can be very useful in many fields, especially for businesses that frequently communicate with customers or other businesses. For example, an auto repair shop needs to provide reports for customers as well as insurance companies. Writing these reports can be automated if supplied with key information such as type of damage, cost, etc. An online retailer needs to create description for many products, which can also be automated given product name, specs and images. In general, people spend a huge amount of time on writing every day, and we hope to make it much easier than today across a wide variety of applications.
Q: How did you get into this space?
A: I got my PhD in computer science in 2007 from University of Illinois with my thesis focusing on applying machine learning algorithms to big data. Then I spent four years working as a researcher at Microsoft applying machine learning to web search. Later I worked at Google and Houzz on various machine learning problems. In recent years, deep learning technology has been growing at an enormous speed so I started exploring how to get it to help our lives in a practical way. I focused more on Natural Language Processing with deep learning, to see how it could automate writing tasks, when we have to write something that we may not be good at. Resume writing is one of them.
Q: What comes next for you?
A: As I am answering your questions, I am still typing out a letter at a time – isn’t this peculiar for an era when automation and artificial intelligence seem capable of disrupting such behavior? For example, drawing has evolved over time; first we used brushes, then we had photography, and along came Photoshop, illustrator, Prisma and all types of AI-powered image generation software. However, when it comes to writing, we still type letters one by one, as if we are unable to revolutionize what typewriters were able to do years ago. Writing resumes is only one application of automated writing by deep learning technology. We will try to automate writing in all fields. The goal is to make it possible for us to write just by providing key information, and have machines control the theme and style of the text being written.