FRIEND.OS: An AI Memory That Remembers What Was Waiting for What
This is a submission for the Hacktoberfest Weekend Challenge: Build for a Friend What I Built I built FRIEND.OS , a local-first AI memory system for a close friend who tends to remember individual things but sometimes forgets the connection between them . The problem is simple: “I’ll do X after Y happens.” We remember X. We remember Y. But when Y finally happens, we may forget that X was waiting…
This submission is for the Hacktoberfest Weekend Challenge: Build for a Friend What I Built. The creator, Kusa L., developed FRIEND.OS, a local-first AI memory system for a close friend who sometimes remembers individual things but forgets the connection between them. FRIEND.OS addresses the problem of forgetting the relationship between two actions, such as “I’ll do X after Y happens.”
For example, the user might remember to submit their internship application after receiving their certificate, but later forget that the certificate is required for the application. FRIEND.OS catches this missing connection and proactively reminds the user to complete the dependent action before attempting the prerequisite action.
The system also handles the opposite situation, where the user is about to perform an action before fulfilling its prerequisite. For instance, if the user says they will eat dessert after finishing rice, but decides to eat dessert now, FRIEND.OS alerts the user to eat rice first. FRIEND.OS doesn't wait for the user to ask what they forgot.
Instead, it detects when something the user has been waiting for happens, or when they attempt to do something before the prerequisite is satisfied. The project's core idea is to store memories, track their relationships, and provide insights based on those connections. FRIEND.OS is built around a real problem experienced by a close friend rather than starting with a generic AI chatbot and finding a use case afterward.
The demo demonstrates two aspects of FRIEND.OS: when something is completed and when the user forgets something. In the first scenario, the user creates a commitment to submit their internship application after receiving their certificate, then receives the certificate, and FRIEND.OS recognizes the dependency and surfaces the next action.
In the second scenario, the user creates a commitment to eat dessert after finishing rice, then decides to eat dessert now, and FRIEND.OS alerts the user to eat rice first. The repository includes a 60-second Demo Mode with three realistic scenarios: internship application, VLSI project report, and research internship. FRIEND.OS is structured as a local web application using Next.js, React, TypeScript, Tailwind CSS, FastAPI, SQLite, and Ollama running Qwen3-4B.
The architecture involves a text browser, Next.js/React/TypeScript, FastAPI, SQLite, Qwen3-4B via Ollama, structured memory, a deterministic dependency engine, and proactive insights.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
