This article offers a glimpse into the dynamic research environment at Megagon Labs, where researchers are pioneering advancements in Natural Language Processing (NLP). Jin Wang’s experience highlights the key advantages of working at a company like Megagon Labs.
What's the most surprising or exciting aspect of working at Megagon?
Working at Megagon has brought me some new challenges as well as opportunities. Here, we can build something from scratch instead of working within an already established pipeline, as happens at big companies. One of the perks of doing research is having outcomes that can be used in real-world scenarios, and with Megagon, that’s really the case. I work on publications, but my work can be applied to the real world, such as improving a commercial recruitment platform. Something else I find intriguing and satisfying is the opportunity to work with real-world data, which is very different from what I saw doing research in academia, where we used benchmark data.
What are you looking forward to with your work now?
I am working towards developing a prototype system for the NL2SQL task, which provides a natural language query interface for relational databases. To reach this goal, I aim to utilize the power of large language models and propose a multi-agent pipeline to improve the proposed system’s usability and performance.
What kind of future do you want to see realized with the technology you are researching?
I want to apply the proposed system to real scenarios to make it easier for users unfamiliar with SQL to query the databases. This could be potentially helpful in applications such as the HR domain, where end-users like recruiters and job seekers might be non-technical users. Besides, the multi-agent pipeline could also help experienced users unfamiliar with the internal logic of the database to quickly formulate their queries and help researchers test their prompt strategies over benchmarking datasets.
How do you keep yourself abreast of what is happening in your field? How does Megagon contribute to how you keep up with current events in your field?
I read the latest published papers in the field. I’ve been keeping up this habit for 5 years. I can always read the papers that have been newly published in the field of data management. I also review papers submitted to top data management, data mining, and artificial intelligence conferences. Because the papers are new, I see new ideas in the drafts or manuscripts.
Megagon Labs helps me keep up with advancements in NLP research. Having summer interns helps us connect to the advisors and their current research work; it’s a way to widen our network within academia. Additionally, Megagon Labs brings in outside speakers. The talks from external contacts at various universities let us know what is happening in related fields and research within academia.
Through the series of invited speakers and my own reading, I have seen the trends and strides made in the research.
What else would you be doing if you weren't in your current career?
I would be a writer and traveler. I like to write ancient Chinese poems. Chinese poets made great literary compositions from beautiful sceneries in the old days. Since there are also so many scenic and natural wonders in the United States, I hope to do the same on my time off. 🙂
Written by: Jin Wang and Megagon Labs
To read more about Jin and Megagon Labs’ work, visit our publications page.