Welcome to ContiAI: Building the Future of Dynamic AI
An introduction to the company, our mission and vision.
Wenjie Huang
CEO & Co-Founder
ContiAI is a Deep Tech company headquartered in Singapore, founded in April 2025. We are dedicated to building the next generation of AI infrastructure equipped with continual learning and memory capabilities, enabling AI to adapt to the world's dynamic changes and evolve from static to dynamic systems.
Our Team
Our team is composed of young researchers and experienced engineers from top universities such as the National University of Singapore, the University of Chicago, and the University of Cambridge, combining cutting-edge research expertise with industrialization experience.
Core Technology
Our core technology is built upon continual learning, with a focus on lifelong model editing, aiming to address the widespread problem of catastrophic forgetting when large models learn new data. Compared to traditional fine-tuning or solutions, our approach offers significant advantages in cost, latency, and stability, allowing AI to learn in real time and reduce memory loss during usage.
Research Progress
On the research side, our team has recently submitted several papers to top AI conferences such as ICLR and AAAI, and open-sourced part of our work. Notably, CleanEdit stabilizes model updates by automatically pruning ineffective past edits, while REPAIR enhances large model performance through intelligent memory management.
Benchmarking Initiative
In parallel, we are building a benchmark for large language model editing, establishing a closed loop across data, models, editing, and evaluation, fostering both academic exchange and product development.
Technology Transfer
On the technology transfer side, we are developing our first API and advancing POCs in multiple industries:
- Intelligent Customer Service: Significantly improve answer accuracy and model personalization
- Finance: Validate the model's ability to adapt instantly to real-time news
- Embodied Intelligence: Explore enabling robots to continually learn and adapt in real-world environments
These cases highlight the strong application potential of ContiAI in high-value industries.
Looking Ahead
Looking ahead, we plan to launch continual learning plugins that can be embedded into mainstream AI workflows, positioning ContiAI as the continual learning engine for all AI systems. The company is currently raising a new round of funding, which will primarily be used for technology R&D, product acceleration, customer expansion, and global market development.