The latest articles on DEV Community by Jinal301409 (@jinal301409). https://dev.to/jinal301409 https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F4154988%2Ff10f8178-8e9f-47e7-9f8a-04730dfcc4db.png https://dev.to/jinal301409 en Jinal301409 Fri, 02 Oct 2026 10:45:52 +0000 https://dev.to/jinal301409/vivabuddy-a-local-ai-examiner-i-built-for-my-friends-dbms-viva-4e7d https://dev.to/jinal301409/vivabuddy-a-local-ai-examiner-i-built-for-my-friends-dbms-viva-4e7d <p><em>This is a submission for the <a href="https://dev.to/challenges/hacktoberfest-weekend-2026-10-01">Hacktoberfest Weekend Challenge: Build for a Friend</a></em></p> <h2> What I Built </h2> <p>VivaBuddy is an AI examiner that runs fully on my laptop. I built it for my friend [Name], who has a [subject] viva on [date]. [Name] prepares by re-reading notes, but freezes when the examiner asks a follow-up question. VivaBuddy asks questions from [Name]'s own notes and adapts: a strong answer gets a harder follow-up, a weak one gets an easier question from another angle.</p> <h2> Demo </h2> <p><a href="https://drive.google.com/file/d/1w60YCTo1ogxUJZ3-YGB6uqfGfTmxyDPL/view?usp=sharing" rel="noopener noreferrer">https://drive.google.com/file/d/1w60YCTo1ogxUJZ3-YGB6uqfGfTmxyDPL/view?usp=sharing</a></p> <h2> Code </h2> <p>GITHUB REPO LINK :- <a href="https://github.com/Jinal301409/Hacktoberfest-vivabuddy.git" rel="noopener noreferrer">https://github.com/Jinal301409/Hacktoberfest-vivabuddy.git</a></p> <h2> How I Built It </h2> <p>Notes are split into sections and embedded locally with <code>nomic-embed-text</code>.</p> <ul> <li>Gemma (<code>gemma3:4b</code>) runs through Ollama and grades each answer only against those notes.</li> <li>Scores are averaged per topic in code, and the model only writes the final recommendation.</li> <li>Stack: React, Node/Express, MongoDB Atlas.</li> </ul> <h2> Why Does Open Innovation Matter? </h2> <p>The model runs locally, so [Name]'s notes are never sent to an AI company.</p> <ul> <li>It costs nothing per question.</li> <li>I could swap models: I started on a 1B Gemma, then moved to 4B when the small one graded too generously.</li> <li>Honest note: MongoDB Atlas is cloud-hosted, so notes and scores are stored there. Only the AI model runs locally.</li> </ul> <h2> My Agent Session </h2> <h2> Prize Categories </h2> devchallenge weekendchallenge hf26challenge