Hacktoberfest Weekend Challenge: Build for a Friend
This is a submission for the Hacktoberfest Weekend Challenge: Build for a Friend What I Built I built DevDraft, an AI-powered communication agent designed for a friend who recently started their first software engineering internship. Like many junior developers, my friend suffers from intense imposter syndrome when communicating in professional engineering channels. They were spending 30–45…
This is a submission for the Hacktoberfest Weekend Challenge: Build for a Friend. The author developed DevDraft, an AI-powered communication agent designed to assist a friend who recently began their first software engineering internship. The intern struggles with imposter syndrome when collaborating with senior engineers, often spending considerable time drafting and editing their communications.
DevDraft alleviates this anxiety by transforming the user's raw notes into polished, professional standup updates or GitHub Pull Request descriptions, eliminating apologetic language and formatting them for Slack or GitHub. The live demo can be accessed at https://devdraft.onrender.com/, although the demo utilizes the Groq API for speed and the app is designed to run locally.
The source code is hosted on GitHub at https://github.com/PalakshiVerma/DevDraft. The architecture was split into a Streamlit frontend and a FastAPI backend, enabling flexible LLM logic. The core of the project utilizes the Meta Llama 3.1 (8B) open-weight model. The FastAPI backend receives the raw text and injects it into system prompts, simulating a confident senior engineer.
The FastAPI backend is deployed as a Web Service on Render, which simplifies the deployment process from GitHub, providing a stable endpoint for the Streamlit UI. The project emphasizes open innovation, offering a "Strictly Local" mode that allows the user's friend to run DevDraft entirely offline using Ollama, without exposing any proprietary company data.
This feature ensures secure, private, and free assistance for junior developers. The submission also aims for the Prize Category Best Use of Render, as the core AI inference API is hosted and deployed using Render's Web Services, bridging the UI and the open-weight LLM.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
This story
This is one outlet's version. Read the fullest account.