i take a lot of notes but rarely find time to organize them so the value of most notes in my personal context quietly disappears.
Notecast is a local note engine i've been building to help me with that. it runs a three stage LLM pipeline (Classify -> Organize -> Consolidate) that automatically builds and maintain a knowledge graph from the notes. the theme hierarchy emerges by subdivision as notes accumulate. Any change generates a proposal that can be edited and commited by the user.
It is early stage and a lot of architectural and domain logic decisions might change but the core is working and is already useful.
It has Obsidian vault integration. I recommend using it (just set vaultPath on configurations)
I'm actively developing it this year and would love feedback.
How is it different from NotebookLM?
Good question,
Notebook LM is a retrieval static tool that answers questions based on the documents you upload to it.
Notecast is an active organizer that takes your notes and continuously builds and evolves the semantic graph around them. It create a DAG of themes to help with navigation and organization. The goal is not to answer questions from the notes but to make them understand their own relation to each other over time within your personal context. For example: if a med student has a lot of medical material, the pipeline, if configured correctly should create the DAG of themes with names relevant to that specific context and by doing that, you have an organized library of your own notes.
NotebookLM is a smart search layer over your material while Notecast is a system that organizes your notes into a "living" structure, closer to a second brain.
Right now it is at early stage so there are a lot of interesting features im thinking about to implement. The most important one(that I think will most differentiate it from NotebookLM and any other available tool) is the practical UX: soon it will have a clean interface that will serve as a quick place to capture, classify, and organize notes with HITL proposals(possible reviewing and editing before applying changes). The core is already there, it just needs some calibration still.
Also, it is local based, BYOK and supports local Ollama models.
Makes sense, local-first is imperative now I guess. Good luck!
thank you!
[dead]
[dead]