Remind AI - Ask about what you’ve seen and done on your laptop

@rayane_melbouci congrats on you launch! This product is amazing!

@rayane_melbouci reMind's open-source approach and focus on privacy are commendable. The integration of AI models for indexing and retrieval offers a powerful solution for managing and accessing digital content. Here are some questions for further exploration: - Can users customize the AI models or indexing methods used by reMind? - How does reMind ensure the privacy and security of user data stored locally? - Is reMind compatible with various operating systems and devices? As a developer with experience in AI and open-source projects, I'm interested in how reMind works. Here are some potential areas for collaboration: - Contributing to the development and refinement of the AI models used for indexing, retrieval, and summarization. - Exploring additional features like tagging, categorization, or integration with other productivity tools. - Fostering a community around reMind to encourage user contributions and support. I'd love to learn more about reMind's technical architecture and explore opportunities to contribute to its development. Additionally, I'm interested in trying out reMind to see how it can improve my productivity and digital organization.

@nitesh_jamod Completely local, free, and open-source.

@kyrylosilin It transforms the data into vectors and stores them in a vector database, which takes up much less space than basic storage. Then, it performs a semantic search to determine the context, identifying the relevant date range. Afterward, it transforms the question into a vector and looks for similarities with the vector store to use that context to answer the questions!

@renchu_song It first performs a semantic search of the question to evaluate the date range and uses that as context by identifying the vector similarities between the question and the context!

@lucasbenjaminfoster Hi Lucas, yes, in a way, it’s possible. This possibility will depend on the model being used. For example, the models supported by Ollama that work in French will be effective for queries in both French and English. To fully support such a feature, you’ll need to use a multilingual model specifically designed for that purpose. I can implement it if you want :)

@danielharrison Let me know if you need some help to implement it :)

@alexandra_describy Thanks !

@cxconnectioncoach As long as it boosts the productivity of many people, with features being added progressively and the project remaining open source, I don't see how it could be negative, even if the project might get overlooked. It's something I created purely for fun, and I’ll continue developing it over time :)