The latest articles on DEV Community by Nowsheen Jahan (@nowsheenjahan). https://dev.to/nowsheenjahan 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%2F3894378%2F1a3f29e9-7624-4c6f-9b9f-2d39521ca67b.png https://dev.to/nowsheenjahan en Nowsheen Jahan Sun, 04 Oct 2026 18:46:45 +0000 https://dev.to/nowsheenjahan/linguapartner-ai-a-lightweight-language-learning-partner-built-for-a-friend-255g https://dev.to/nowsheenjahan/linguapartner-ai-a-lightweight-language-learning-partner-built-for-a-friend-255g <h2> What I Built </h2> <p>I built <strong>LinguaPartner AI</strong>, a lightweight, full-stack conversational language-learning platform designed specifically for a close friend who wanted a low-pressure, judge-free environment to practice foreign language skills daily without the anxiety of traditional classroom feedback. </p> <p>Instead of an intimidating UI, the app acts as a friendly digital partner that engages in natural dialogue, keeps track of conversational flow, and helps correct grammar mistakes on the fly in a supportive, conversational manner.</p> <h2> Demo </h2> <ul> <li> <strong>Live App URL:</strong> <a href="https://lingua-partner.onrender.com/" rel="noopener noreferrer">https://lingua-partner.onrender.com/</a> </li> <li> <strong>GitHub Repository:</strong> <a href="https://github.com/nowsheen19jahan/lingua-partner.git" rel="noopener noreferrer">LinguaPartner AI Repo</a> </li> </ul> <h2> How I Built It </h2> <p>The application is architected around a robust, asynchronous <strong>FastAPI</strong> backend built with Python. </p> <ul> <li> <strong>Request Validation:</strong> Utilized <strong>Pydantic</strong> to handle strict schema validation for incoming chat payloads, target language flags, and response structures.</li> <li> <strong>Architecture:</strong> Designed following clean modular REST principles to ensure low latency and seamless communication between the frontend client and the backend services.</li> <li> <strong>Deployment:</strong> Hosted live on <strong>Render</strong> using a containerized Python environment configured for optimal performance on free-tier infrastructure.</li> </ul> <h2> Why Does Open Innovation Matter? </h2> <p>Building with an open and self-hostable backend ensures total privacy for chat logs and user practice data. It eliminates reliance on costly, closed-source enterprise APIs, allowing developers to build free, customized tools tailored specifically for the people they care about.</p> <h2> My Agent Session </h2> <p>I used an agentic workflow to rapidly structure, scaffold, and refine the application architecture, helping design clean routing patterns and streamline the overall project setup from scratch.</p> <h2> Prize Categories </h2> <ul> <li>Best Use of Render</li> </ul> devchallenge weekendchallenge hf26challenge