Software engineer Emily Stacy built a portfolio that doubles as a learning journal, using Google Cloud Run for deployment and Google Gemini for AI-powered summarization. The project, submitted to Google's 'New Year, New You' challenge, showcases a full-stack application with a React frontend and Node.js backend, designed to capture technical reasoning over time.

The system's core feature is a custom summarization endpoint that sends journal content to gemini-2.5-flash. It includes a resilient extraction layer to handle inconsistent AI outputs, a practical approach to real-world API limitations. This treats AI as a helper rather than a source of truth, with graceful fallbacks when summaries can't be confidently generated.

Stacy prioritized a learning-first design, moving beyond flashy demos to solve the personal problem of retaining complex knowledge. The portfolio demonstrates full-stack ownership, from frontend UI to Cloud Run deployment labels. It represents a practical application of AI for developer productivity, focusing on structured reflection over simple project showcases.