# Echoes: Bringing Memories Back to Life with Open-Source AI > Source: > Published: 2026-10-30 19:06:46+00:00 I built **Echoes**, an AI-powered memory preservation and storytelling application inspired by the theme "Ghost in the Machine." The idea is simple: memories and personal stories deserve to be preserved. Echoes aims to help people turn personal memories and stories into meaningful written narratives using open-source AI. The application uses Python and Streamlit for its interface and an open-weight Gemma model through Ollama for local AI inference. The AI workflow is designed to transform user-provided memories into structured stories while preserving the original information. The application is built around local inference rather than depending on a proprietary hosted AI API. Open innovation gives developers greater control over how AI applications are built and operated. Using open-weight models and local inference makes it possible to process personal stories on a user's own machine without sending their content to a third-party hosted AI service. It also makes model experimentation, customization, and independent development more accessible. For a memory-preservation application, privacy and control over personal stories are especially important. Best Use of Gemma — include only if Gemma is meaningfully integrated and the category requirements are satisfied. The past should not disappear just because time moves forward. With open-source AI, we can build tools that help preserve stories and memories while keeping users in control of their data.