This video, Create an AI Handwritten Task Evaluator with Gemini API and Apps Script, demonstrates how to build an AI chatbot mini-app using Google Apps Script and the Gemini API to evaluate students’ handwritten assignments against a predefined rubric.

Key Features:

– Handwritten Assignment Evaluation: Students can upload assignments as an image or PDF. The AI extracts the handwritten text (demonstrated with Thai handwriting) and evaluates it against instructions and grading criteria [00:24 Opens in a new window ].

– Rubric-Aligned Feedback: Instead of open-ended generation, the bot delivers structured feedback highlighting strengths and specific areas to improve prior to final submission [00:46 Opens in a new window ].

– Student Record Retrieval: Students can save their evaluation results with a custom passcode to review their feedback history later [01:03 Opens in a new window ].

– Automated Instructor Logging: Submissions, student details, and AI-generated feedback are automatically saved to a Google Sheet [01:09 Opens in a new window ].

– 5 Steps to Build the Application Google Doc Setup: Create a document that contains the assignment prompt and grading rubric [01:32 Opens in a new window ].

– Google Sheet Setup: Set up a spreadsheet to act as the database for student information and evaluation logs [01:45 Opens in a new window ].

– Gemini API Key: Obtain an API key via Google AI Studio to connect the grading logic to the model [02:01 Opens in a new window ].

– Code Generation amp; Configuration: Generate code.gs and index.html (e.g., using Gemini Canvas) [02:08 Opens in a new window ].

– In code.gs, configure environment variables including your Gemini API key, Google Doc ID, and Google Sheet ID [02:17 Opens in a new window ].

– The index.html file creates the chat interface, upload controls, and base64 file conversion [02:30 Opens in a new window ].

– Deployment: Deploy the Apps Script project as a Web App to get a shareable URL for students [02:39 Opens in a new window ].