Roadmap
What features and characteristics are planned for future versions of Remembrances
OpenSource and Free Version
It will always remain free and open source under MIT license.
- SurrealDB support in server and embedded mode
- Embeddings support (ollama and OpenAI embeddings API)
- Local GGUF models support for embedding generation
- GPU acceleration support (Metal, CUDA, ROCm)
- Knowledge base support (markdown files)
- Tools support similar to Mem0 but in a single binary without scripting language dependencies like Python or NodeJS, faster and more efficient.
- Short-term immediate memory support (key-value)
- [In progress] Support for source code and software projects indexing using code-specialized embeddings, implementation of tools inspired by Serena but with faster and more efficient indexing with AST and Tree-Sitter.
- AI agent for complex data ingestion and understanding how to store it through MCP sampling
- AI agent for deep knowledge querying through MCP sampling
- Implementation of knowledge reinforcement algorithms, generalization, and selective forgetting
- Support for advanced knowledge files (PDF, DOCX, etc)
- Support for multimedia files (images, audio, video) with multimodal embeddings
- Support for time series
Commercial Version
- Support for multiple users and work teams
- Support for advanced auditing and logging
- Support for automatic backups and restoration and migration/export between database instances (both embedded and external SurrealDB server)
- Support for enterprise integrations (LDAP, SSO, etc)
- Web administration and monitoring interface
- Knowledge visualization interface (graphs, statistics, etc) and integrated chatbots for direct querying of stored knowledge