At HubX, our primary focus has always been—and will continue to be—creating user-friendly products that make the latest AI models accessible to end users. In line with this commitment, we are expanding our efforts by entering the generative AI research and development space. This expansion includes establishing an independent AI Research team at the İzmir Institute of Technology.

The lab will be directed by experienced professionals from Silicon Valley who have held positions at prominent companies like Apple, Google, and Meta. These experts boast over a decade of proficiency as ML Engineers and Researchers, with a demonstrated history of successfully leading large-scale AI projects. They will be instrumental in developing HubX's AI strategy and leading major research initiatives. Additionally, we will work in close partnership with professors from the İzmir Institute of Technology, whose extensive knowledge of AI and ML fundamentals will greatly enhance our research activities and help cultivate the upcoming generation of AI experts.

The AI Lab operates with a clear mission—to build new base models and fine-tune existing foundation models for specific use cases in our current and upcoming apps.

Our AI-powered apps have reached more than 80 million users in the past year. Every day, people discover new ways to utilize generative AI for various purposes. By analyzing the behaviors of millions of users and the specific reasons why and how they engage with these models, we have identified many intriguing “niche” use cases.

While there are excellent open-source base models like Stable Diffusion, Mistral, and Meta’s Llama for generic use cases, and APIs like OpenAI's GPT-4, these models often lack the necessary modifications for specific applications or can only be modified to a certain extent. Our research is focused on training models specifically for unique use cases to solve real-world problems more effectively.

Training a foundation model or fine-tuning an existing model is a time- and resource-intensive endeavor, requiring a dedicated and experienced research team.

We are committed and excited to take on this challenge. We will be sharing updates on our research progress, so stay tuned.

Share this story