DopikAI’s automated machine learning (AutoML) framework helps users build AI models quickly and efficiently. Automating the end-to-end machine learning process offers the advantages of simpler and faster production of solutions and models that often outperform hand-designed ones. Our common technologies used in AutoML include AI Training and System Design.

Technologies: DeepLearning, Tensorflow, Pytorch, Onnx, Docker, K8S, AWS, Python, NLP, Computer Vision
Duration: 6 months
Team size: 7

Problem
The trend of integrating AI in business is becoming more popular thanks to its performance and cost-effectiveness. Companies in need of AI technology have spent a lot of money, time and effort recruiting resources to implement these tools, whether the need is simple or complex. In development, companies will have to pass sensitive datas to developers. Therefore, the need for a customizable solution that can help companies build AI tools quickly, efficiently, safely and customizable is very high.

Solution
DopikAI’s AutoML framework utilize AI training and system design. The system allows people without in-depth AI knowledge to create highly customizable AI models quickly and accurately. With the support of advanced technology, companies can test or improve their business processes at the lowest cost, while the quality remains competitive.

The system is designed in two parts: AI framework and UI system. The first part is the AI framework running on Tensorflow or Pytorch, which has easy-to-use functions and is easily monitored and cloud-ready. The second part is an UI system where users can design AI models by drag-and-drop. The system also includes intensive tests guaranteeing professional and safe improvements and maintenances.