Research Has More Shape Now
Arivu now turns notes and sources into typed research objects, imports documents and transcripts as searchable notes, keeps recent knowledge readable offline, and preserves more structure in backups.
Saved knowledge gets more useful when it has shape.
A bookmark is a source. A note is context. A task is an open loop. But real research also has projects, people, books, meetings, decisions, and threads that unfold over time. Those things show up again and again, but they do not always fit cleanly into a tag or a folder.
This update adds that missing layer. Arivu can now turn saved material into typed research objects, bring documents and transcripts into the same note workflow, trace topics across time, and keep recent knowledge readable when the network drops.
Objects for the Things You Keep Naming
Some ideas keep coming back with the same name.
A launch project. A person you are researching. A book you keep quoting. A meeting that created follow-ups. A decision you do not want to lose. A research thread that spans weeks of notes and sources.
Arivu now gives those things a lightweight home as knowledge objects. The first object types are:
- projects
- people
- books
- meetings
- decisions
- research threads
An object can have a title, a description, small structured fields, and an optional source item. That source might be a bookmark, a note, or another object. The point is not to create a heavy database or a project management system. The point is to give important recurring things a stable shape.
What this means for you: A decision can be more than a paragraph buried in a daily note. A meeting can be more than a calendar event. A research thread can be more than a tag. These things become first-class enough to find, connect, and preserve.
Documents and Transcripts Join the Same Loop
Not everything worth saving is a web page.
Research often arrives as a PDF, an EPUB, a Markdown file, a pasted transcript, a text export, an HTML document, or OCR copied from an image. Before this update, those materials had to be converted into some other shape before they fit naturally inside Arivu.
Now document and media-style imports can become normal searchable notes. Upload or paste the material, and Arivu stores the result as a note with source context. From there, it behaves like the rest of your second brain:
- it can enter Inbox
- it can be searched
- it can be linked to bookmarks or notes
- it can carry tasks and reminders
- it can appear in Review
- it can be exported and backed up
Image text extraction can use a configured provider when available, but the workflow does not require every import to become an AI job. Pasted OCR text and transcript text are useful on their own.
What this means for you: A saved article, a PDF handout, a video transcript, and a meeting note can all live in the same workflow instead of becoming separate piles.
Topic Evolution Turns Time Into Context
Some topics are not single items. They evolve.
You might mention a product direction in a daily note, save three sources about it, write a meeting note, record a decision, then come back two weeks later with new evidence. Search can find each piece. Topic evolution shows how the thread moved over time.
Arivu can now line up a topic across daily notes, saved pages, standalone notes, meetings, decisions, and knowledge objects. That is useful when the question is not only “where is the source?” but “how did this idea change?”
There is also a fixed Today board that gathers the working state into a simple view: Inbox, Working, Review, recent decisions, and recent meetings. It is not an infinite canvas. It is a practical board for the shape Arivu already understands.
What this means for you: Your second brain can show a thread as it unfolds, not just individual search hits.
Calendar Imports Become Meeting Objects
Meetings are often where research turns into commitments.
Arivu can now import calendar event text and create meeting objects from it. A meeting object can then sit beside the notes, decisions, reminders, and source items it relates to.
This matters because meetings often create the next action, but the context usually lives somewhere else. The calendar knows when the meeting happened. The notes know what you thought. The sources know why it mattered. Arivu is starting to keep those pieces closer together.
Offline Recall Gets More Practical
Self-hosted tools should still feel usable when a network connection is rough.
Arivu now keeps bounded local snapshots for recent read-only second-brain surfaces after successful online reads. That includes places like Today, saved pages, notes, Review, Inbox and work queues, reminders, memory jogger, and typed search.
If the connection drops later, those surfaces can show the latest local copy instead of failing immediately. Writes still need the server, except for dashboard URL captures that can queue locally and replay when the signed-in browser is online again.
What this means for you: A flaky connection does not have to turn your recent second brain into a blank page. You can keep reading recent context, then write when the server is reachable.
Quotes Can Point Back to Their Source
Annotations are more useful when they remember where they came from.
Reader annotations now preserve enough source context to jump back to the matching passage when the archived page still contains it. That makes quotes feel less like copied fragments and more like anchored references.
What this means for you: Highlighted passages stay connected to the source text instead of becoming detached clippings.
A Narrow Door for Local Automation
There is also a new path for people who want local automation around Arivu.
CLI-audience routes can support scoped search, reading saved bookmarks, notes, and object context, creating notes, adding tasks or reminders, and recording decisions. This is not a broad autonomous assistant with unchecked write access. It is a narrow HTTP surface for controlled local clients that need to work with your Arivu instance.
That direction fits the same principle as the governed assistant: automation should be useful because it is scoped, inspectable, and boring.
Other Changes
- The installed app now opens to Today by default, matching the daily operating model.
- Full JSON backups now preserve knowledge objects and restore their source references under the importing account.
- Document imports, object workflows, Today board, and topic evolution are reflected in the self-hosted app without adding a separate frontend dependency tree.
Arivu started by making saved links searchable. Then it added the loop: Today, Inbox, Focus, Review. This update gives that loop more shape. Sources can become notes. Notes can become objects. Objects can show up in decisions, meetings, boards, and topic timelines. The archive is still there, but it is less flat now.
That is the direction: not more storage for its own sake, but a self-hosted second brain that understands the forms research actually takes.