Teachers are building their own classroom software to solve everyday teaching challenges without writing traditional code. By using artificial intelligence to build applications through plain-language commands, a process known as "vibe coding," educators can bypass expensive, rigid commercial software. A recent classroom initiative shows how this approach connects students with physical books already sitting on school library shelves.
What Happened
David Webb, an international school teacher with 11 years of experience, spent months designing an AI-powered reading recommendation system to connect students with underused libraries. Instead of hiring software engineers, Webb described what he wanted to build to large language models in plain English. Despite setbacks, including a lack of backups that wiped out hours of work on his 12-year-old computer, Webb launched his app, LibraryAid, in November 2025. Webb initially used Microsoft Copilot to generate code but switched to Claude to resolve frequent errors. The tool allows teachers to upload their existing school library catalogs as CSV files. The AI then processes student reading levels and classroom topics to generate custom reading lists from the school's actual inventory.
The Bigger Picture
This local software development matches findings from recent academic research. A study published in Frontiers in Education showed that AI-adapted reading materials improved reading comprehension for lower-proficiency students while keeping advanced students engaged. Other academic tools, like T-TExTS, use knowledge graphs to recommend literature, which reduces teacher workloads.
However, custom AI tools raise concerns about student data privacy. Under the Family Educational Rights and Privacy Act (FERPA), any software handling student names or reading progress must comply with strict federal rules, as noted by Beni Education. Standard Data Privacy Agreements (DPAs) often fail to protect students because student data is still sent directly to third-party cloud servers for processing, a vulnerability explained by ibl.ai. This risk has driven demand for comprehensive district evaluations, a trend we previously highlighted in our coverage of urban school board mandates.
What This Means for Families
For parents, these custom apps offer tailored support for their children. A child who struggles to find books may receive accurate, personalized lists pulled straight from their school library. However, because individual teachers build these tools in their spare time, the software often lacks the security reviews required for commercial edtech. Parents must weigh the educational benefits against the privacy risks of third-party data processing.
What You Can Do
- Ask school library staff how they track and recommend books, and if they use automated systems.
- Ask school administrators how they review teacher-built AI tools for FERPA compliance before classroom use.
- Request that any classroom tool processing your child's reading records or personal information operates under strict cloud-privacy standards.