As usual, here is our internal product report for June 2025 
.
We’ve had 4 releases with 625 value points (440 → 400 → 625 last 3 months).
Main product areas and how they will change with time in near future.
Permissions
Capability-based Space permissions
We are (finally!) in the process of migrating to the new capability-based mechanism for Space access. At this point, nothing changes for the users, it’s purely an enabler for our next features. Okay, perhaps, Fibery will work slightly faster, especially for large workspaces, once we fully migrate.
Most internal services already support the new mechanism, you can see the progress based on the Dev Tasks here: Allow not to extend access from Space to its Databases.
The first two meaningful things after the migration would be:
- Allow to have Databases outside Spaces (in Default Space)
- Consolidate default access template names across Space, Database, and Entity levels
Access audit
We’ve introduced access audit on Space and Database levels so that architects can tell who has what final access and why:
- Show all Users who have access to a Space
- Show all Users who have access to a Database
Next step: Introduce access audit for Users and User Groups from the other side.
Adoption
Bad news first: it doesn’t look like those who have created but haven’t used custom access templates started to do so after Explain access templates:
It’s no better for workspaces created after our changes went live (Apr 24):
Maybe, people are just exploring custom access templates 
. We’ll need a proper qualitative research (aka a few interviews) to figure out what’s going on.
Good news: 40% of all collaborative workspaces now utilize User Groups and automatic access via People Field while over 30% share Databases, both numbers higher than expected:
Import
CSV
Over 200 workspaces have used CSV import in the last 30 days which shows the demand, and quite a few customers have complained about its deficiencies which shows the room for improvement. So we’ll continue working on the CSV import till we (and customers) are largely satisfied.
These were the two main things we delivered in June:
- Allow non-Admins to import data into existing Database
- Make switching from new to existing Field in integrations Field mapping discoverable
And here are the next major candidates:
- Import self-related entities via a single CSV file
- Update existing entities via CSV import
- Introduce optional Field settings in integrations and import
Additionally, it should be possible to import via the AI workspace builder.
Discoverability
A few small things:
- Suggest import or sync when adding a new Space
- Replace “Import data” with a nested “Import from…” menu
- Update headers for Database screens, including adding
Importbutton.
We’ve postponed Skip authentication method choice if OAuth is available for the connector until we find someone with Vizydrop expertise to join the import team.
Notion
After we’ve learned how to fetch rich-text from Notion, the next step is to Sync Notion standalone Pages to Fibery Database.
Plus, make the connector stable.
Duplicates
Now architects can Automatically merge duplicates when enabling unique toggle:
The functionality is fully ready now, so we’ve started advertising it with the New badges.
If there is enough demand, we’ll introduce a way to find and merge duplicates without requiring future values to be unique.
Home page
We’ve released My Work widget and got immediate backlash in the community, since it is not flexible enough. This feedback came from mostly creators and it is unclear about its usage by mere mortals. We also did not promote it and did not make Home Page default (so far).
Then we had a conversation and decided to… implement Dashboard View. Plan is this:
- Embed Views into Home Page with a dashboard-like layout
- Basic Dashboard View (it will support Context Dashboard View as well as a result)
- Admins want to configure Home Page for other users
So Kostya will spend July in these things. Behold!
Gantt View
Gantt View was released for Pro and Enterprise plans. It immediately became more popular than Map View and catching up with Thread View now 
.
We’ve made a serious mistake and based our Gantt View on board model, which lead to inability to show nested self-relations… Now we are re-doing Gantt View to support List model, fixing some bugs and polishing it. Should be done in ~2 weeks.
Then we will at least announce it in-product and see whether usage will be better.
AI
We’ve released the first version of AI Build mode as experimental, and some accounts started to use it. Overall, usage is not so great, but it is growing, so there is hope.
AI Agent got new tools, like semantic search and Fibery user guide, so now it is more capable.
Build mode has serious problems, since it can create databases with corrupted schema. We hope to fix these problems in the first week of July and release new getting started experience in July.
Other things we are working on:
- Add all views to Build mode (now it supports Table and Board only)
- Generate rules and buttons
- Make UI more pleasant with better context handling and artefacts
Whiteboard
In May, we released the first section updates and in June we kept improving them. You can now link and lock sections, and working with objects inside them is smoother. We also started improving how groups and selected objects behave. To better preserve data, we began updating the Undo history and will continue with Activity and Version History for whiteboards.
What we’re working on now and planning next:
- Groups improvements
- Use arrow keys to move multiple selected items
- Make “Undo/Redo” work well
- Whiteboard version history and activity
- Embed views into the whiteboard (reports, tables, etc)
P.S. Your feedback is welcome! 
