The latest articles on DEV Community by Guo X (@guo_x_36a154b5e3620014fc8). https://dev.to/guo_x_36a154b5e3620014fc8 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%2F4024588%2F4ec6ad3e-636d-4f8c-8534-e71ac47bcdcd.png https://dev.to/guo_x_36a154b5e3620014fc8 en Guo X Sun, 12 Jul 2026 12:17:37 +0000 https://dev.to/guo_x_36a154b5e3620014fc8/heirloom-ai-preserve-family-memory-46m3 https://dev.to/guo_x_36a154b5e3620014fc8/heirloom-ai-preserve-family-memory-46m3 <p><em>This is a submission for <a href="https://dev.to/challenges/weekend-2026-07-09">Weekend Challenge: Passion Edition</a></em></p> <h2> What I Built </h2> <p>Heirloom AI preserves family recipes &amp; skills using Gemini multimodal AI. Upload handwritten cards, cooking photos, voice recordings, or gesture videos — it generates structured archive entries with poetic memory cards, evidence-based ingredients (with confidence levels), physical-cue step guides, and prominently flagged "speculative gaps" for family verification. Includes AI illustration generation and image enhancement. Full-stack React + Express app, persists in localStorage, exports Markdown.</p> <h2> Demo </h2> <div class="crayons-card c-embed text-styles text-styles--secondary"> <div class="c-embed__content"> <div class="c-embed__body flex items-center justify-between"> <a href="https://heirloom-ai.ai.studio" rel="noopener noreferrer" class="c-link fw-bold flex items-center"> <span class="mr-2">heirloom-ai.ai.studio</span> </a> </div> </div> </div> <h2> Code </h2> <div class="ltag-github-readme-tag"> <div class="readme-overview"> <h2> <img src="https://assets.dev.to/assets/github-logo-5a155e1f9a670af7944dd5e12375bc76ed542ea80224905ecaf878b9157cdefc.svg" alt="GitHub logo"> <a href="https://github.com/gxobst" rel="noopener noreferrer"> gxobst </a> / <a href="https://github.com/gxobst/Heirloom-AI" rel="noopener noreferrer"> Heirloom-AI </a> </h2> <h3> Transform messy family recipe cards, verbal kitchen instructions, or raw cooking photos into beautifully structured, editable archive entries with Gemini. </h3> </div> <div class="ltag-github-body"> <div id="readme" class="md"><div class="markdown-heading"> <h1 class="heading-element">Heirloom AI 🌾</h1> </div> <blockquote> <p><strong>Preserve the recipes, rituals, and skills that live in family memory.</strong></p> </blockquote> <p>Heirloom AI is a web application designed to transform messy personal materials—such as raw photographs, chaotic scribbles, handwritten notes, verbal stories, videos, and voice recordings—into beautiful, structured, editable, and shareable archival entries.</p> <p>The first MVP focuses heavily on <strong>family recipes</strong>, because culinary traditions are deeply emotional, practical, highly visual, and uniquely susceptible to being lost across generations.</p> <div class="markdown-heading"> <h2 class="heading-element">📖 What It Does</h2> </div> <p>Rather than generating generic, standardized recipes off the web, <strong>Heirloom AI</strong> acts as a warm oral-history and preservation assistant. It processes your specific memory context and images to draft custom archives containing:</p> <ol> <li> <strong>Evocative Memory Cards</strong>: A warm, poetic summary and quote summarizing the tradition.</li> <li> <strong>Personal Narrative &amp; Lore</strong>: Captures the emotional voice and regional background.</li> <li> <strong>Evidence-based Ingredients</strong>: Checklists tracking what ingredients are visible or described.</li> <li> <strong>Physical Cue Guides</strong>: Instruction steps…</li> </ol></div> </div> <div class="gh-btn-container"><a class="gh-btn" href="https://github.com/gxobst/Heirloom-AI" rel="noopener noreferrer">View on GitHub</a></div> </div> <h2> How I Built It </h2> <p>Single Node server running Express + Vite SPA (no CORS issues). Two Gemini models: gemini-3.5-flash with schema-constrained JSON (responseSchema) for structured archive generation, and gemini-3.1-flash-lite-image for watercolor illustrations + image editing. Dynamic prompt enhancement based on uploaded file types (video→gesture analysis, audio→oral history, document→OCR). Client-side image compression (800px max). localStorage with graceful degradation (strips non-image data first, then all data URLs if quota exceeded). Confidence levels are a feature — uncertain items get animated amber alerts with suggested family follow-up questions. Editorial brutalism design with warm earth tones reinforcing the scrapbook preservation feel.</p> <h2> Prize Categories </h2> <p>Google AI</p> devchallenge weekendchallenge