Background
Shenzhen Landray Software Co., Ltd. (Landray Software) has been a pioneer in the digital office field since 2001. As a nationally recognized high-tech enterprise, a contributor to the national standard for Knowledge Management, and a top 10 Xinchuang supplier, Landray Software's mission is to "make organizations smarter". The company provides comprehensive digital transformation support for organizations of all sizes through products such as the Landray MK Digital Work Platform and the aiKM Intelligent Knowledge Management Platform. Its services cover a range of digital solutions. These include Platform as a Service (PaaS) platforms for large and medium-sized enterprises, Office Automation (OA) collaborative office systems, enterprise portals, Business Process Management (BPM) flows, low-code platforms, Knowledge Management, smart contracts, and Xinchuang office solutions. These offerings help customers accelerate their digital transformation.
Development challenges and solutions
During its digital transformation, the Landray Software R&D team faced several challenges. Continuous product iterations led to a massive codebase, which placed significant pressure on the development team.
- Time-consuming code comprehension: The large codebase, accumulated over many product versions, made it difficult for developers to understand existing logic. They spent excessive time deciphering code, which slowed down development.
- Lack of timely optimization suggestions: While developing new features, the team lacked prompt code optimization suggestions. This often resulted in logic errors and poor code quality. Additionally, manual code reviews added significant costs to the development process.
- High training costs for new employees: New employees spent considerable time studying existing project code to learn the company's coding standards. The team's well-established standards and codebases were not effectively leveraged for training, which increased onboarding costs.
- Efficient assisted coding: Qoder CN uses a model that is optimized with a specialized code corpus and fine-tuned for popular programming languages such as Java, Python, Objective-C, JavaScript, and TypeScript. It combines this with a project-wide awareness of related code across files to significantly improve coding accuracy. This technology enhances development efficiency and reduces rework caused by coding errors.
- Code explanation: The Qoder CN code explanation feature helps developers and maintenance teams quickly understand code logic and identify areas for optimization. This feature significantly reduces code review time and improves maintenance efficiency.
- Code optimization: For new feature code, Qoder CN quickly identifies potential issues, acting as a rapid code review. This helps resolve problems during the coding stage, avoiding major changes later and saving time and resources.
- Adapting to custom development scenarios: Qoder CN allows enterprises to customize and extend prompt tasks. It integrates with an enterprise's private data to provide retrieval-augmented code completion and R&D Q&A, meeting specific development needs and improving R&D flexibility.
- Security and compliance: Qoder CN allows enterprises to customize policies for filtering sensitive information. It pre-filters information, such as passwords and emails, from code snippets at the plugin level. End-to-end data encryption and protection help enterprises establish a secure and compliant development environment.