星球科技 builds an education-focused AI platform centered on exam paper digitization and automated grading, powered by LangChain, LlamaIndex, and ChatGPT. The tech stack reveals a full-stack education play: backend services (Python, C#, ASP.NET Core, SQL Server, Redis), frontend UI (Vue, uni-app, Bootstrap), and cloud infrastructure (Alibaba Cloud). The project list—exam paper import, OCR optimization, AI agent design, and micro-lesson content—indicates a system designed to automate the administrative and instructional burden on teachers and schools.
星球科技 operates a platform that digitizes and automates workflows around examination and course delivery in K-12 and higher education. Core capabilities span exam paper onlineization (OCR and content tagging), AI-driven grading (rule-based and model-based), course production, and video content management. The platform targets institutional customers—schools and exam boards—rather than individual learners. The small, focused team is split evenly between engineering and specialized functions (education, product, research, sales), reflecting a product-market fit phase rather than early exploration.
LangChain, LlamaIndex, and ChatGPT form the core AI stack. The platform uses these to power automated grading, Q&A optimization, and AI agent design for educational workflows.
Alibaba Cloud and Alibaba Cloud OSS for hosting and object storage. Backend services use Python, C#, ASP.NET Core, SQL Server, and Redis; frontend uses Vue and uni-app.
OCR accuracy for exam paper digitization, prompt optimization for Q&A systems, and exam paper creation process optimization are documented pain points in active development.
Other companies in the same industry, closest in size