Megagon Team Profile: Eser Kandogan

As Head of Engineering, Eser helps further develop the skills of the engineers on the team while also helping drive the direction of the lab. He came to us with a wide range of computer science and mentoring experience. Read on to learn more about Eser and get further insight on becoming a successful engineer and mentor.

How many years of CS experience do you have and do you remember what first sparked your interest in CS?

I started programming in middle school. I had a ZX Spectrum with 48K RAM, essentially a basic gaming computer back then. I always liked computer games; some days I would easily spend 8 or 10 hours playing. I would only play during summer breaks though. Back then, there wasn’t too much concern about screen time, I guess. My parents were happy to see me using my computer. They figured it was teaching me something, and it did, as that was also the beginning of my journey with computer programming. 

I remember one summer my computer broke, and so I wrote my programs on paper. I would debug them on paper, too. I recall using millimetric paper to design my game characters. As a kid, I would also write letters to gaming companies in the US to get hired as a computer programmer!

My formal education in CS was years later. I studied Computer Science and Engineering in Turkey, at Bilkent University, where I earned a scholarship to attend the program. In 1994, I came to the United States for my PhD, at the University of Maryland, College Park. Upon completion of my PhD in 1998, I moved to California to work at Silicon Graphics (SGI). SGI’s main business was computer graphics, and I loved computer graphics. My background in data visualization and analytics fit right into the data mining group I belonged to at SGI. I was lucky to work with the people who developed the field’s first data mining libraries. We worked on three-dimensional visualizations of data mining algorithms like association viewers, clusterings, and such. 

In 2000, I joined IBM Research. There, I worked with many great people who made fundamental contributions to computer science, from data management systems to storage systems, text analytics, visualization, and of course, human-computer interaction. Beyond CS, I also learned a lot from my colleagues from different fields like Cognitive Science. Since then, I spent several years doing (amateur) ethnography to understand how large computer systems at data centers are administered. That pursuit opened my eyes to how critical it is to study people and computers in a holistic manner, as socio-technical systems. In 2019, I joined Megagon Labs. 

So, it all started with computer games. But today, the computer games I play are very simple like slither.io.

What advice would you give current engineers striving to get to a leadership position? What should they do outside of work to improve?

For anyone in engineering, I’d say read more, and every day. Over and over, I find myself doing two types of reading. One is technical blogs. They are great— you get an idea of a topic, algorithm, or approach in a digestible form and in a very short time. They easily fill those idle gaps in your day. If I see a topic coming up more and more, I would buy an actual book on the topic to get more detail. That’s the second type of reading I’d encourage. I like learning from books more than blogs, but however you want to do it is fine so long as you do it. The main point is to try to keep up to date. CS is so fast; everything changes after five years. 

For someone with people-management responsibilities, I’d suggest really getting to know the people you work with, learning about their interests, and encouraging them to lean into their interests so they can succeed. I also encourage the people I manage to try things— go to tutorials and implement so they can really learn it. Learning through practice can go a long way. Going beyond current frameworks and libraries, I encourage my employees to learn the practices, like UI design vs UI development. In all, everyone should be trying new things so they continue to grow and develop.

What sparked your move from a research to an engineering role at Megagon Labs?

I was getting increasingly more interested in Natural Language Processing (NLP) and deep learning. Megagon Labs is doing cutting-edge research at the intersection of NLP, data management, and Human-Computer Interaction (HCI). The cross-disciplinary research topics at Megagon Labs are very appealing for me. Besides that, I wanted to do more product work. I wanted to do more engineering at the forefront. And as a researcher, I was also interested in pushing boundaries, especially by translating research technology to production.

What has been the most exciting project you have worked on or overseen at Megagon Labs thus far and why?

I am really excited about our new project. We are building a knowledge base for human resources. I see it as a valuable project that can have multiple benefits and also introduces new research directions. In essence, it is a resource that can solve many pre-existing problems in job matching and content generation. On the research side, I believe it will enable us to build high-precision solutions that are more data driven. In a way, it will complement superb capabilities of language models with the introduction of transformer models, with data that is specific to the domain, as well as data organized in a manner that exploits properties and relationships, such as a knowledge graph. We expect that combining the natural-language understanding capabilities of language models with data from knowledge bases will help improve precision. Through learned representations from knowledge graphs, we can provide a language for explanations and have better control of the output, whether it is for natural language generation or other tasks like matching.

The project has a strong NLP piece and a strong database piece. All that is interesting to me. 

Tell us about any hobbies or activities you like to do outside of work?

I like spending time with my family and gardening, ideally together. In the past, I used to do a lot of woodworking. I used to build wooden model ships and somewhere along the line, I went into doing woodwork with bigger pieces for the house and yard. I guess I have less time to do high-precision work required for smaller models. Once I finish building my workshop, and the kids get a little older, I am hoping to get back to ship model building, and hopefully get them interested too. 

I also like learning new languages. Five minutes here and there really adds up! For something like learning a language, daily repetition is key. I started learning Thai during the pandemic. I speak Turkish by heritage and studied German for seven years, plus I learned English during college. I must say I also had many failures, Japanese and Mandarin being two.

As the manager of the Engineering team and with an integral role in the management of Megagon Labs in Mountain View, in what direction do you want the engineering team to go?

For this question, I would probably focus more on the types of work we should be doing more than on a technical direction. As members of the engineering team at an industrial research lab, we should always do ‘research engineering,’ be embedded in research both in the lab and outside of it, try and get first-hand experience with cutting-edge technology, and examine those technologies from the perspective of practical application. That way, we can see and understand the gap between theory and practice. In my view, a thorough understanding of these gaps will lay the groundwork for further research and development. As research engineers, we should always focus on ‘value’ that is brought by research.

We hope you’ve enjoyed this special interview with Eser Kandogan! Check out our blog to see other team member profiles and learn more about Megagon’s recent work!

Follow us on LinkedIn and Twitter for to stay up to date with us.

Written by Eser Kandogan and Megagon Labs.

Share:

More Blog Posts: