GPU Cloud Solutions Optimized for LLM Research and Development

SELECTSTAR

Selectstar, an AI startup, ranks first in Korea for text data building. In this interview, we spoke with Minwoo Kim, the AI NLP team lead, and Heondeuk Lee, an AI researcher at Selectstar. They shared how they came to adopt Elice Cloud, their achievements, and their future goals.

Q: Hello. Please introduce yourselves and tell us a bit about Selectstar.

Heondeuk Lee : Hello. I’m Heondeuk Lee, an AI researcher at Selectstar.

Minwoo Kim : Hello. I’m Minwoo Kim, the AI NLP team lead at Selectstar. Selectstar is the top company in Korea for text data building, and we are also developing data evaluation solutions. With the rise of large language models (LLM), we are working on a product that utilizes LLM internally, including a character-based video calling solution. We have IP contracts with well-known character companies such as Pororo and Robocar Poli.

Q: How did you hear about Elice Cloud? And why did you decide to adopt it?

Minwoo Kim : I knew Elice as a company that started as an education platform with a great LXP. We were using another cloud service, but while comparing options for research, I learned about Elice Cloud from our CEO. We chose Elice Cloud because it was easy to start and stop instances and was more cost-effective than other options.

Q: What was your first impression of Elice Cloud?

Minwoo Kim : Elice is known for its well-developed LXP, and the GPU cloud was equally user-friendly from a UI/UX perspective. Other cloud services often require numerous settings and terminal operations, but Elice Cloud is easy to set up with its no-code interface.

Q: What are you using Elice Cloud for, and what are your goals?

Minwoo Kim : We primarily use Elice Cloud for research. As the demand for GPUs increases with the LLM era, Selectstar uses Elice Cloud for internal LLM research and development. Our main goal is to strengthen our research base through Elice Cloud, as successful research is crucial for developing new platforms or selling our technology to other companies.

Q: What benefits have you experienced with Elice Cloud?

Heondeuk Lee : It’s very useful to easily track my teammates’ usage. Also, splitting instances and checking our code is more convenient with Elice Cloud compared to using tools like Git. Additionally, managing CUDA versioning is hassle-free.

Q: Can you share any achievements you’ve had with Elice Cloud?

Heondeuk Lee : During the development of a new solution, we needed to complete an experiment within three days and proceed to statistical analysis. By using Elice Cloud to divide the instances and run the experiment in parallel, we quickly and efficiently obtained significant results.

Q: What do you like about Elice Cloud?

Heondeuk Lee : The low cost compared to other GPU services is a major advantage. Also, the ease of dividing and managing instances is a big plus. We can easily collaborate, as team members can see what code has been modified in parallel.

Q: Would you recommend Elice Cloud to other organizations? If so, why?

Heondeuk Lee : I would recommend Elice Cloud to any company that needs to move quickly through the proof of concept (PoC) or minimum viable product (MVP) stages. It’s fast, stable, and you can focus on your work without worrying about lag or version control issues. Elice Cloud is an excellent choice for model development.

Q: What are Selectstar’s future plans?

Minwoo Kim : Selectstar is currently the top company for text data building in Korea, and we aim to expand globally. We have two main goals: to develop a cost-effective, high-efficiency data building solution and to create a SaaS platform that can evaluate LLM performance for different companies. We look forward to achieving these goals and continuing our collaboration with Elice Cloud.

Elice Cloud has been instrumental in our research and development, providing fast, reliable performance and easy management. We look forward to its role in Selectstar’s journey to the global market.

Elice Cloud offers a dedicated AI data center with high-end GPU resources. If you’re a startup looking for a GPU cloud for AI research, consider Elice Cloud!

👉 Learn more about Elice Cloud On-Demand

👉 Get an Elice Cloud GPU service consultation

*Elice owns the copyright to this content, which is protected under copyright law.

*Without prior consent, secondary processing and commercial use of the content are prohibited.