I Built StudyNest to Make Scattered Study Notes Easier to Review

I created StudyNest, a compact Vietnamese study aid designed to help users review disorganized notes. The tool enables you to save notes within a browser, pose queries based on those notes, produce five practice questions along with their suggested answers, or obtain a concise summary. I developed this application during the Hacktoberfest Weekend Challenge.

To evaluate its functionality, I utilized a Python lesson note: when I inquired about the reason behind ket_qua being 34, the application explained that the append(10) function had transformed the list [7, 8, 9] into [7, 8, 9, 10], resulting in a sum of 34.

StudyNest harnesses the open-weight openai/gpt-oss-20b model via Groq's API. The model can be swapped in a local configuration file without altering the study process, allowing the app to operate without downloading a substantial model onto a computer with limited storage space. The application only transmits notes to Groq when an AI feature is engaged; otherwise, notes are archived in the browser's localStorage.

The application's frontend is composed of a single HTML file. A compact Python HTTP server relays the file and interfaces with Groq using the official Python SDK. The server inquires the model to respond exclusively from the supplied notes and to indicate when the notes lack an answer. The API key remains in config.py, which Git ignores; config_example.py illustrates the configuration format.

Having tested the prototype on a machine with constrained disk space, I discovered that moving inference to Groq rendered the application functional on that system while preserving an open-weight model as its foundation. An upcoming version will enable exporting notes and empowering learners to answer quiz questions interactively.