Your brain is non-linear. Your notes shouldn’t be linear either. AltaNote is an AI-powered thinking canvas where ideas connect themselves. Capture thoughts by voice, Telegram, web links, or files. Drop them onto an infinite canvas. AltaNote automatically builds connections, surfaces meaning, and helps you see the whole picture. Ask AI to analyze a note or an entire project — and get structured insights, summaries, and new ideas back. Stop organizing. Start thinking.
I want to make the app a powerful personal assistant. Helpers in the implementation of ideas. A person thinks and creates, an assistant helps and promotes
No reviews yetBe the first to leave a review for AltaNote — Notes That Think Back
Maker
📌
Hi Product Hunt 👋
I built AltaNote because my notes never stayed where my thoughts actually were.
Ideas came as voice messages, Telegram notes, links, screenshots — scattered everywhere. I tried folders, tags, systems… none of them worked long-term. My brain is non-linear, but all note apps forced me to think in lines and lists.
So I built a canvas where thoughts can live the way they appear — connected, messy, evolving. AltaNote captures ideas by voice, Telegram, or web and automatically builds connections between them.
The turning point was adding AI — not to “summarize notes”, but to think back. You can ask AI to analyze a single note or an entire project and get structured insights, summaries, and even new notes back.
This is my attempt to build a second brain that actually works with you.
I’d love your feedback — especially what feels confusing or unnecessary. Happy to answer any questions!
Report
The non-linear canvas approach resonates—I've had the same problem with Obsidian's folder structure not matching how ideas actually connect. The Telegram integration is smart for capturing fleeting thoughts. My biggest question is about the AI connection-building: does it use semantic similarity to link notes, or does it analyze actual content relationships? And when you ask AI to "analyze a project," how does it determine project scope—manual selection, or does it automatically identify clusters?
Report
Maker
@easytoolsdev When building connections, the neural network studies all the elements and builds a relationship based on the meaning of the elements. One condition is that the connections - go from the older elements to the younger ones (by the time of creation), otherwise the canvas turned into a grid. The project scope can be set in the project settings. When building relationships, the model uses it as the main vector.
On the scale of the project, if not to say directly which nodes are needed, it will proceed from the context of the project. The agent receives the notes and tags along with the attachments. Next, he tries to build a system. If something didn't go according to plan the first time, you can repeat the request specifying what is expected.
Locally without encryption, interaction with the server is implemented with encryption.
Currently, work is underway to store data in encrypted form. Note content, attachments, and user data. The feature will be released in March.
Report
Maker
@daniele_packard I would like to add that in the plans for a future release there is an application for the ability to work only locally without accessing the server. But the work has not started yet. We plan to release it in May.
The non-linear canvas approach resonates—I've had the same problem with Obsidian's folder structure not matching how ideas actually connect. The Telegram integration is smart for capturing fleeting thoughts. My biggest question is about the AI connection-building: does it use semantic similarity to link notes, or does it analyze actual content relationships? And when you ask AI to "analyze a project," how does it determine project scope—manual selection, or does it automatically identify clusters?
@easytoolsdev
When building connections, the neural network studies all the elements and builds a relationship based on the meaning of the elements.
One condition is that the connections - go from the older elements to the younger ones (by the time of creation), otherwise the canvas turned into a grid.
The project scope can be set in the project settings. When building relationships, the model uses it as the main vector.
@easytoolsdev
On the scale of the project, if not to say directly which nodes are needed, it will proceed from the context of the project. The agent receives the notes and tags along with the attachments. Next, he tries to build a system. If something didn't go according to plan the first time, you can repeat the request specifying what is expected.
ChatPal
Very cool - how does privacy work? Stays local?
@daniele_packard
Locally without encryption, interaction with the server is implemented with encryption.
Currently, work is underway to store data in encrypted form. Note content, attachments, and user data. The feature will be released in March.
@daniele_packard I would like to add that in the plans for a future release there is an application for the ability to work only locally without accessing the server. But the work has not started yet. We plan to release it in May.
@daniele_packard
If absolute confidentiality is needed, then access to the server must be blocked. But in this case, you lose touch with telegram and the AI assistant.