Case of Intelligent Platform for Mobile Terminal Development
- In this project, Farben provides the customer with a series of innovative measures, including the development coding assistant, the demand collection robot, the R&D knowledge Q&A assistant, and other sub-items, and builds a large model-based intelligent support platform for mobile terminal development. The project solves many problems, including difficulties in repeated function search across different provinces, high costs of gateway API learning, low efficiency of historical code and file retrieval, and insufficient project code learning efficiency. It not only improves development efficiency and code quality but also strengthens teamwork and knowledge sharing capabilities. The building of a unified intelligent development platform delivers automation and intelligence in code development, high efficiency in demand collection, and structuring in knowledge management, laying a solid foundation for promoting the intelligent transformation of the customer’s R&D process.
Case Background
- Waste of resources caused by repeated function development: Projects in different provinces involve massive similar functions, while there is a lack of a quick search and reuse mechanism, resulting in redundant development workload.
Steep gateway API learning curves: There are a great variety of gateway APIs with dispersed files, and it takes massive time and costs for team members to learn and adapt.
Difficult to quickly retrieve historical codes and files: There is no tool for the unified management and efficient retrieval of massive historical codes and files, making it difficult for developers to quickly find the information needed.
Low efficiency of project code learning: When new team members join or in the event of cross-team collaboration, there are barriers to understanding project codes, affecting the overall R&D efficiency.
Case Content
- 1. Coding Assistant：
The coding assistant substantially improves development efficiency and code quality by providing intelligent programming support through large models, covering real-time code continuation, automatic annotations, logic explanations, error correction, and unit test generation. In addition, by combining with the customer’s business code fine-tuning and search optimization, it makes sure to address general programming issues and handle challenges in specific business scenarios to a certain extent. It supports comprehensive and intelligent support for the development team. - 2. Demand Collection Robot
Using the natural language processing capabilities of large models, the demand collection robot performs the hosting of communication records of multiple technical support groups and the intelligent collection of demand through WeChat group chat management. The robot is able to automatically recognize and categorize key demand content, generate a clear list of tasks, and support efficient claim and distribution. By improving the team demand management process, it substantially improves demand sorting efficiency and the accuracy of task execution, providing powerful support for distributed team collaboration. - 3. R&D Knowledge Q&A Assistant
The R&D knowledge Q&A assistant, with a unified knowledge repository management platform at its core, supports the quick retrieval of massive historical codes and files and offers a variety of functions, including intelligent Q&As, natural language search, and file summarization. Users may quickly retrieve the relevant answers from the knowledge repository through the Q&A assistant and invoke basic large model capabilities to complete knowledge Q&As, which fully improves the team’s abilities for knowledge sharing and reuse.
Value for Customers
- Improve development efficiency and quality: The support of an intelligent assistant tool enables the quick search and reuse of repeated functions while improving the efficiency and quality of code writing, optimization, and testing.
- Lower learning costs: The precise search and intelligent Q&As offered by the R&D knowledge Q&A assistant allow team members to quickly master key knowledge in gateway API, project codes, and R&D processes, substantially reducing their learning cycles and time costs.
- Efficient demand management: The automated collection and sorting of demand by the intelligent robot reduces the complicated steps of manual management and considerably improves the efficiency of team collaboration.
- Stronger abilities of knowledge sharing and reuse: A unified knowledge repository management platform provides powerful support for the quick retrieval and reuse of historical codes and files, improving knowledge accumulation and sharing efficiency.
- Promote the intelligent transformation of the R&D process: A unified intelligent development support platform is built to fully improve team efficiency and collaboration abilities, providing a solid foundation for the intelligent upgrading of R&D.
