ByteDance
Internal AI-assisted software development with TRAE
ByteDance said more than 80% of its engineers used TRAE to assist development
Problem
ByteDance sought to reduce software-development work through AI assistance within engineers' coding environment.
Approach
ByteDance developed TRAE, an AI coding environment with code completion and natural-language programming. Technology vice president Hong Dingkun reported widespread internal use in June 2025. He said a substantial share of code was AI-generated, without giving a company-wide percentage.
Results
Takeaway
The reported adoption rate established broad engineering use. It did not quantify development-time savings or changes in software quality.
Categories
- Size
- Large enterprise
- Industry
- Technology and telecom
- Business function
- IT and R&D
- Innovation type
- Process innovation
- AI technology
- Generative AI
- Adoption
- Built in-house
- Evidence
- Company's own claim
Sources
Source summaries
Beijing Business Today reported Hong Dingkun's statement that more than 80% of ByteDance engineers used TRAE. It described code completion and natural-language programming features, without reporting measured internal productivity gains. [Beijing Business Today report](https://www.bbtnews.com.cn/2025/0612/559772.shtml).
Full details are in the original sources.