Get a DemoStart Free TrialSign In

Interview

5 min read

For the next interview in our series speaking to tech founders from around the world, we’ve welcomed Seung Oh, Co-Founder and CEO of Data B, the company behind Engram, the first AI-powered writing platform designed for non-native English speakers.

Contents

Tell us a bit more about the business you represent.

Data B is the developer of Engram, an AI-powered grammar checker and paraphraser optimized for non-native English speakers. Our proprietary AI, combined with ChatGPT, provides sophisticated, context-based edits for more natural-sounding text.

Unlike other proofreading or paraphrasing services, Engram is uniquely optimized for non-native speakers. We've refined our algorithm to catch mistakes that other grammar checkers usually miss, like awkward word order and phrasing. In addition, instead of offering individual word corrections, Engram offers sentence-level suggestions to create text that sounds more natural.

Can you share a little bit about yourself and how you became a Founder?

Prior to co-founding Data B, I led product management at Amazon and Samsung Electronics. As a Senior Product Manager at Amazon in 2016, I was constantly writing six-pagers to articulate what needed to be built. The other PMs and I used to joke that one of Amazon’s Leadership Principles that read “leaders are right a lot” actually meant “leaders write a lot.”

My manager once reviewed my proposal and put corrections all over the paper in red, saying "Your writing is choppy. Study grammar." I felt really embarrassed at first, but then I thought, I might not be alone. There are millions of people whose native language isn't English facing challenges while studying and working in English. This became the problem I wanted to solve.

I tried many grammar checkers back then, but none of them were really helpful. I was so frustrated by this because people all over the world were talking about AI and machine learning, and yet no one had developed software that could solve my problem. Therefore, I ended up co-founding my own company and developing Engram, an AI-powered English writing assistant for non-native speakers.

Being the Founder, what do your day-to-day responsibilities look like?

My main responsibilities include product management, marketing, sales, and other administrative work. I spend approximately half of my day in product management, a quarter in marketing, and the remaining quarter in sales and other administrative tasks.

For product management, I am involved in the entire product development process from defining requirements to testing. I occasionally write proposals for new products or features, answer questions from the engineering team, and revise requirements when unnecessary implementation challenges arise. I participate in testing once the development of a new product or feature improvement is completed.

For marketing, I set directions for our marketing efforts and assign responsibilities to individual members in the marketing team. I check the progress of our marketing efforts and make adjustments to increase the speed of our growth.

For sales, I communicate with both current and prospective customers, addressing their inquiries. I identify and define requirements for products in order to better meet their needs.

Other administrative tasks involve managing payroll, preparing reports for investors, completing tax and government subsidy forms, drafting and reviewing employment contracts, and paying bills.

How have you evolved as a Founder and leader over time, and what have been some of the biggest lessons you've learned along the way?

One of the biggest lessons I have learned along the way is that I should have been communicating with our customers more frequently. We recently conducted a series of interviews with our paying customers which provided valuable insight into how they use our product and the challenges they encounter along the way. For instance, they wanted to use a translator in addition to our proofreader since they frequently sought assistance with translating documents that were originally written in Korean. I was surprised because I thought that people who could construct English sentences on their own would rarely need translators. Another interesting finding was their desire to see suggestions presented in various tones. That was surprising because I thought that having too many options would actually make it harder to choose one. I wished I had done these interviews earlier. If we had conducted these interviews a few months earlier, we would have completed these features by now.

What is your company's approach to innovation, and how do you foster a culture of creativity and experimentation among your team?

Whenever someone comes up with a new idea, I initiate discussions to explore the reasoning behind it, rather than relying on my intuition to judge it. If there are reasons to believe that the idea might work, my team complements it with additional research and detailed plans for implementation. If disagreements arise, we devise strategies to minimize our investment while implementing metrics to evaluate the impact of those ideas. In addition, I continually generate new ideas to maintain our growth momentum even if the initial ideas are unsuccessful. This approach helps the team propose ideas without putting too much effort on persuading others, resulting in a culture of creativity and experimentation.

The approach is founded on the realization that my intuition can often be inaccurate. I've seen many cases where some ideas looked bad at first sight, but turned out to be a huge success. For example, I first thought that Airbnb would not work because people would be afraid of staying at someone else's home. Likewise, some of the ideas that came from my co-founder appeared unimpressive at first glance but proved to be remarkable when we implemented them. These humble experiences have taught me the importance of being cautious when evaluating others' ideas based solely on my own intuition.

The recent emergence of generative AIs like ChatGPT, supported by large language models, has become one of the biggest trends. Large language models offer unprecedented capabilities that can help improve the performance of our proofreading and paraphrasing algorithms, thereby enhancing the quality of our service. However, they also pose challenges for us to compete against both general-purpose services like ChatGPT from big tech companies and specialized competitors, such as Quillbot.

We strive to stay ahead of the competition by focusing on finding the best technologies that can accomplish our goal of helping non-native English speakers. Thanks to ChatGPT, many companies are releasing their large language models either via API or as open source. Our engineering team invests a significant amount of time and energy into conducting research on these models. We also test them with our own data and undergo an evaluation process to choose the best-performing ones for our services.

What can we hope to see from your business in the future?

We aspire to provide everything that non-native English speakers need for effective communication in English. We aim to expand our reach to all non-native English speakers worldwide. With the suite of products that we offer, they will be able to elevate their English to the next level.

Can you tell us what book you are reading currently?

I am currently reading Elon Musk by Walter Isaacson. This is my second book by Walter Isaacson, following Steve Jobs. The book provides me with valuable lessons in business and management. It also teaches me how to face challenges and solve problems with determination.

If you've enjoyed this article why not read our Interview with Founder Reeve Benaron or What is Infrastructure Monitoring next?

Get the latest elastic Stack & logging resources when you subscribe

© 2024 Logit.io Ltd, All rights reserved.