Interview with Seth Killian, Founder of Lingco Language Labs
Seth Killian, Founder of Lingco Labs
By Shannon Spasova, University of Michigan
Shannon Spasova: Hi, can you tell us a little about yourself?
Seth Killian: I’m Seth, the founder of Lingco Language Labs. I’m an undergraduate student at Michigan State majoring in Computer Science and minoring in French. I love traveling to new places and pushing the edge of what’s possible with technology. My primary interest is in educational technology and how we can we can use data analysis and artificial intelligence to deliver personalized learning.
Shannon Spasova: How did you become a language learner? When did you start learning languages? What appeals to you about learning languages?
Seth Killian: Admittedly, I initially became a language learner in high school by requirement rather than by choice. It wasn’t until I studied abroad that I became passionate about and could really appreciate the benefits of learning another language. I find learning additional languages appealing because it expands my perspective and opens up new parts of the world. When I go abroad and can speak the language, I feel like I can experience and take part in the culture rather than be a passive observer.
Shannon Spasova: Why did you choose to become a computer science major?
Seth Killian: When I was 11, my parents brought home a new computer. Curious how it worked, I took a screwdriver and completely dismantled it. Luckily, I was able to reassemble it. When my grandfather found out, he was impressed and bought me a couple of computers that were sold for parts at the Michigan State University Surplus Store. I was able to fix and sell them on Craigslist. I took the money from the first two and bought four more. Then I sold those and nine months later had a little over 140 computers in my grandfather’s basement. At the height of the endeavor, he called me one morning asking why there was a UPS truck in his driveway with a pallet of computers from the University of Virginia that I purchased wholesale. In order to keep track of the inventory, I built a database program and soon after became more and more interested in software. When it came time to pick a major, computer science seemed to be the logical choice.
Computer science is interesting to me because I have all of the tools that I need to build a program on my computer. For many engineering majors, it takes years and significant resources to build something impactful whereas Mark Zuckerberg was able to build Facebook on a personal computer from his dorm room in only a couple of weeks. Since then, the pace of technology has only increased which is why I find computer science to be so exciting.
Shannon Spasova: Can you tell us about your company?
Seth Killian: Lingco is an intelligent language learning platform for schools. It predicts what parts of a language students already know and recommends assignments to instructors based on what their students should learn next.
Instructors are able to use our platform to create vocabulary sets and activities. With vocabulary sets, instructors can choose or create lists of vocabulary for their students to learn. We look at the history of each learner which is comprised of all of the questions that they’ve answered and the collective history of everyone on our platform along with features such as the lexical similarity and frequency. With this information we can predict how likely a student will be able to comprehend and produce a word, which we then use to create personalized study sessions.
With activities, we enable instructors to easily create engaging content. Instructors can do things such as ask a question by recording themselves and students can respond with a recording which can be automatically graded with our voice recognition system. They can combine this with other prompts and content such as videos, multiple choice questions, drag and drop matching, recordings, etc.
Shannon Spasova: What made you decide to create your own company, and how did you manage to merge your two interests of technology and languages?
Seth Killian: Back in high school, I was an awful French student. After three years, the only things that I could say were “hello” and “today is my birthday”. The summer after sophomore year of college, I wanted to study abroad. When I asked my parents, they said absolutely not, reminding me that I wasn’t a very strong French student to say the least. I was able to talk them into a deal: if I could learn French in three months, they would send me abroad. They figured that it was a pretty safe bet since I wasn’t able to make much progress in three years.
I tried a lot of online resources, but they were frustrating because they didn’t take any of my previous knowledge and exposure into account. They just asked if I was a beginner, intermediate, or advanced student and placed me into a level. The problem for me was that I had a vague idea of what I knew, no idea of the mistakes I made, what I needed to work on next, or go back and review. Learning a language isn’t a linear process and even though I knew a lot of “intermediate” aspects of the language, there were still parts from the beginner level that I didn’t remember or never learned in the first place. Luckily, I was able to find a graduate student who was finishing up her Ph.D. and was willing to tutor me. Working with her, every week I was able to speak a little more French and my parents became a little more nervous. Thanks to her help, in the end, they had no choice but to send me!
While in Paris, it was clear to me that I was there because of the personalized help from my tutor. She was able to design a personalized lesson plan for me based on exactly what I knew and needed to work on next. Of course, this process isn’t very scalable, which is where I thought that technology could step in.
Computers are great at analyzing lots of data and making predictions. My idea was if we could use data to predict what parts of a language students knew, we could predict what they’re ready to work on next and give them a personalized experience. I began building the first version of Lingco while still in Paris for my own use and when I returned, I began to turn it into a platform for the classroom.
Languages and technology are two things that I’m passionate about and starting a company with Lingco allows me to do what I love and to build a platform that countless other students and instructors can benefit from.
Shannon Spasova: What is your process like in improving the product that you are working on?
Seth Killian: I think that a platform meant to be used by students and teachers should be built by students and teachers. This idea is intuitive, but we see products all too often that are designed for an end user by a bunch of engineers who have never actually sat down with the user and asked what is and isn’t important.
We spend a lot of time engaging with our users. We have a feedback page where instructors can add new ideas and vote and comment on other ones. Our team uses this to drive the roadmap and development of the platform. Having instructors work with us also makes it easy for us because rather than guessing what they want to see in Lingco, they just tell us!
The added advantage of Lingco being a cloud-based software is it that allows us to get features into the hands of our users in days rather than months. Historically, software would be released every six to twelve months, but now we’re able to release updates as soon as they’re finished. On an average day, the platform will be updated between 2 and 3 times.
Shannon Spasova: What is unique about your company or your product?
Seth Killian: Most platforms take a one size fits all approach. Every student answers the same questions regardless of their individual level. Lingco uses an adaptive engine that is predictive and efficient.
For students, this means less time reviewing what is already known and allows for them to focus on material needing more attention because Lingco is able to individualize content and feedback to the learner. For instructors, the benefit is that Lingco shows real-time snapshots of where students have deficits. Instructors see what students know and Lingco recommends how they can bridge the gap between the proficiency level and course goals. Traditional classroom assessments only tell an instructor after the fact what they should have spent more time on.
With Lingco, every time a student answers a question, its overall picture becomes more accurate. As students move up course levels, Lingco stays with them so the next instructor has the benefit of all of the previous data and doesn’t need to start from scratch.
Shannon Spasova: Some states have started allowing coding to count to fulfill a foreign language requirement. As a person who is a coder but also a language learner, do you agree with this? What do coding and language learning have in common, and what is different about them?
Seth Killian: Both computer and human languages are similar in that they’re systems for communication, as are all languages. The difference is that a computer language uses a context-free grammar and a human language uses a context-sensitive grammar.
Unlike humans, computers are unable to think for themselves. So as a result, a computer language must be defined by a context-free grammar, which is a finite set of rules on how to interpret a statement. Since there’s only one way to interpret a statement, there’s no nuance. This means that a program with a thousand lines of code and a single misplaced comma won’t be able to compile and run – it either works completely, or not at all. Human languages on the other hand are context-sensitive, so the meaning of a word is dependent on the other words in the sentence and its placement relative to them. So, a misunderstood word in a conversation won’t result in a total breakdown in communication.
Although foreign and human languages have similarities and I can understand how they would be grouped together, the outcomes of learning them are different. Learning to code is valuable for developing critical thinking, logic, and problem-solving skills, much like learning algebra. Learning a language, on the other hand, provides not only the clear outcome of being able to communicate with more people, but I think that learning another language and all of its nuances provides a new perspective and broadens the understanding and appreciation of the world around us.
Both disciplines are valuable, but I don’t think it makes sense for one to replace the other because they have different learning outcomes and this would leave a gap. If anything, I see computer science as a more appropriate replacement for an upper-level math course.
Shannon Spasova: What are your future plans for your company and your career?
Seth Killian: After graduating from Michigan State, I look forward to taking a full-time role at Lingco where I plan to be for the foreseeable future. I’m excited to continue to work with instructors to create a best in class language learning platform. I see a lot of significant changes in the near future to educational technology and how we use it to learn languages. This is due to the emergence of virtual reality and the increased availability of cloud computing resources. I’m looking forward to being a part of this and pushing the bounds of what we think is possible.
Is there anything else that you would like us to know about you or your company?
Seth Killian: If any readers are interested in trying Lingco Classroom, my email is email@example.com – I’d be happy to provide a demo of the platform and set up a demo account.