The latest articles on DEV Community by Imran Hossain (@imran_hossain_c5416804ee8). https://dev.to/imran_hossain_c5416804ee8 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%2F4192202%2F62320002-6999-4a72-8a80-d870510d5997.png https://dev.to/imran_hossain_c5416804ee8 en Imran Hossain Sun, 18 Oct 2026 16:55:42 +0000 https://dev.to/imran_hossain_c5416804ee8/hacktoberfest-weekend-challenge-build-for-a-friend-4639 https://dev.to/imran_hossain_c5416804ee8/hacktoberfest-weekend-challenge-build-for-a-friend-4639 <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><strong>Moye</strong> is a patient, distraction-free learning companion built specifically for young kids who find traditional educational apps frustrating. It features zero timers, no stressful scores that drop, and a friendly mascot named Moyin to guide children through simple questions one at a time.</p> <h2> Demo </h2> <ul> <li> <strong>Live Application:</strong> <a href="https://moye-bay.vercel.app" rel="noopener noreferrer">https://moye-bay.vercel.app</a> </li> <li> <strong>Video Demo:</strong> <a href="https://youtu.be/Hp7FgtXbPpc" rel="noopener noreferrer">https://youtu.be/Hp7FgtXbPpc</a> </li> </ul> <h2> Code </h2> <ul> <li> <strong>GitHub Repository:</strong> <a href="https://github.com/ayodejiades/moye" rel="noopener noreferrer">https://github.com/ayodejiades/moye</a> </li> <li> <em>License:</em> Open-source under the MIT license so other developers and educators can inspect and build upon it.</li> </ul> <h2> How I Built It </h2> <p>Moye is built using a modern, lightweight web architecture:</p> <ul> <li> <strong>Framework:</strong> Next.js 16, React 19, TypeScript, and Tailwind CSS</li> <li> <strong>Core Logic:</strong> Uses Bayesian Knowledge Tracing (<code>lib/mastery.ts</code>) to dynamically adjust question difficulty based on a child's responses without relying on heavy cloud servers.</li> <li> <strong>Accessibility:</strong> Built with a custom comfort panel supporting OpenDyslexic font, custom letter spacing, and speech synthesis across multiple languages.</li> </ul> <h2> Why Does Open Innovation Matter? </h2> <p>Open innovation was vital for this project because a child's learning attempts and potential struggles are highly sensitive data. Relying on a closed, commercial API would mean sending private user telemetry to a third-party server. By utilizing open-source libraries, web standards, and local storage, the application runs entirely client-side, ensuring complete privacy, zero per-user API costs, and full offline functionality.</p> <h2> My Agent Session </h2> <ul> <li> <strong>DevRelay Session:</strong> Built using iterative AI agent workflows with local validation tests ensuring complete reliability.</li> </ul> <h2> Prize Categories </h2> <ul> <li><strong>Best Use of Open-Source AI</strong></li> <li><strong>Best Personal Productivity / Learning App</strong></li> </ul> devchallenge weekendchallenge hf26challenge Imran Hossain Sun, 18 Oct 2026 15:12:18 +0000 https://dev.to/imran_hossain_c5416804ee8/touch-grass-mission-2nmd https://dev.to/imran_hossain_c5416804ee8/touch-grass-mission-2nmd <p><em>This is a submission for the <a href="https://dev.to/challenges/hacktoberfest-week1-2026-10-05">Hacktoberfest Open-Source AI Challenge Week 1: Touch Grass</a>.</em></p> <h2> What I Built </h2> <p><strong>NatureLens AI</strong> is an open-source outdoor companion designed to get people off their screens and into the natural world. Rather than keeping users glued to endless feeds, NatureLens acts as a quick, frictionless bridge to the outdoors: you snap a photo of a bird, plant, or trail landmark, and the app instantly identifies it, gives you a fascinating ecological fact, and then <strong>tells you to put your phone away and look up.</strong> </p> <ul> <li> <strong>Who it’s for:</strong> Hikers, curious park-goers, students, remote workers needing a mindful break, and anyone who wants to use technology to enhance real-world exploration rather than escape into virtual ones.</li> </ul> <h2> Demo </h2> <ul> <li> <strong>Live Web Application:</strong> <a href="https://hacktoberfest2026-week1.vercel.app/" rel="noopener noreferrer">https://hacktoberfest2026-week1.vercel.app/</a> </li> <li> <em>Tip:</em> Try uploading a sample outdoor photo or test the local exploration mode to experience how it guides you outside quickly.</li> </ul> <h2> Code </h2> <ul> <li> <strong>GitHub Repository:</strong> <a href="https://github.com/sujitkr268/Hacktoberfest2026_week1" rel="noopener noreferrer">https://github.com/sujitkr268/Hacktoberfest2026_week1</a> </li> <li> <em>License:</em> Open-source under the MIT license so developers can freely fork, inspect, and build upon it.</li> </ul> <h2> How I Built It </h2> <p>NatureLens AI is built around a modern, fast, and privacy-conscious stack:</p> <ul> <li> <strong>Framework:</strong> Next.js 16 with App Router</li> <li> <strong>Frontend &amp; Styling:</strong> React 19, TypeScript, and Tailwind CSS v4</li> <li> <strong>Open-Source AI Integration:</strong> Built to interface with open-weight vision models (such as local execution via Ollama or hosted open-weight models via Groq API endpoints). Server-side Next.js Route Handlers securely process prompts without exposing keys on the client.</li> <li> <strong>Local-First Privacy:</strong> Because the core identification engine can run locally via open-weight models, users' outdoor photos and location data never have to be logged on a centralized corporate cloud.</li> </ul> <h2> Why Does Open Innovation Matter? </h2> <p>Open innovation is essential for a project focused on nature, privacy, and unplugging. Relying on a closed, commercial API would mean forcing users to stream their personal outdoor moments, trail paths, and private photos directly to third-party corporate servers. </p> <p>By leveraging open-weight models and local inference tools (like Ollama), NatureLens guarantees complete user data privacy. Furthermore, open innovation allows developers worldwide to customize the model weights for regional flora and fauna—making local conservation data accessible anywhere, even deep on a trail with zero cellular network signal.</p> <h2> My Agent Session </h2> <ul> <li> <strong>DevRelay Session:</strong> Embedded and tracked via open development logs to demonstrate iterative prompt tuning for concise, distraction-free UI design.</li> </ul> <h2> Prize Categories </h2> <ul> <li> <strong>Best Use of Gemma</strong> (or open-weight integration)</li> <li> <strong>Best Use of DigitalOcean / Render</strong> (hosted deployment runtime)</li> </ul> devchallenge hf26challenge