AI Platform
(AI PRODUCT ENGINEERING)
The AI Product Engineering group at Torus is in charge of 3 main domains:
- IT Infrastructure of Torus (high-performance servers, local cloud network, collaborative tools, cyber security, infrastructure and process optimization, etc.)
- Software & System Engineering (APIs, mobile apps, etc.)
- AI Platform & AI Engineering (data management system, data annotation system, training system on the cloud, servicing system, tools for constructing and fine-tuning AI models, etc.)
The group serves the computing and AI software engineering needs of the other groups within the company. Together with our partners, we also offer AI development and consulting services for the other companies, based on our AI Platform.
We stay ahead of the curve by constantly updating our knowledge base and developing our skills based on the latest research and advancements in the fields of machine learning and deep learning. The cornerstone of our work is not only the creation of AI models but also the development of efficient systems, infrastructure, and processes to optimize AI development.
Our meticulous approach to system and process design ensures an optimized, streamlined, and effective workflow, enhancing the speed and quality of our AI development projects.
Technology Stack
Our flexible and diverse technology stack guarantees that we can develop robust and efficient AI solutions that meet our clients’ unique needs. We constantly explore and adopt new technologies to stay at the forefront of AI development.
-
Machine Learning & Deep Learning Frameworks
We use state-of-the-art frameworks like TensorFlow, PyTorch, and Keras to design, train, and deploy our AI models.
-
Programming Languages
Python is our primary language for AI development, highly favored for its simplicity and wealth of machine learning libraries. We also use languages such as C# and ReactJS for specific tasks and system development.
-
Cloud Platforms
We utilize AWS and Azure for production deployment and rely on Kubernetes on-premises for AI training and internal system management.
-
AI Development Tools
To streamline our workflow and ensure the efficiency of our AI development process, we employ tools like Kubeflow, MLflow, FiftyOne, and others.
-
AI Platform
Our team has developed a bespoke AI Platform, tailoring it to optimally support our specific needs and projects.
-
DevOps & Infrastructure Tools
Docker and Kubernetes aid us in achieving seamless integration and deployment.
Partners
