🐖 Fibery Product Report: January 2026

In January we did 4 releases with total 480 value points (600 → 530 → 820 → 480 last 4 months). Releases quality and speed is decent in January. And December was very good.

Main product areas and how they will change with time in near future.

Permissions

Since Andrew was busy with validation rules, we didn’t start any Features in January — only finished and released Manage custom access templates for Database access from UI.

Adoption: guests (retraction)

In November, I wrote that Guests have seen incredible adoption among collaborative workspaces. I was wrong, and my incredulous intuition was right. There was a bug in the query that only affected Guests and only in this chart. Here’s the reality:

So 17% became 9% and 68% became 13%. Yikes!

More reasons to consider Improve Guests user experience! But not before…

Next: access to users

The next major thing is Configure access to Users, and the roadmap is clear now:

It’s not the fastest way to first value, but I believe it’s the right path for the long term.

Once the access to users is ready, there will be two major things left to close the Granular Access theme:

  • Allow to hide Space data (entities) from Admins
  • Add create capability to access templates

Validation rules

Initially we wanted to implement access to previous values in automation rules, so we dig into it. However, it appeared that validation rules are is a better way to solve data consistency problemm is a be need mostly, so we jumped into Validation rules with brave Andrew and fearless Sergey. We almost completed all the things, the only remaining use case should be released soon Support Previous field values in validation rules.

After than Andrew will rest for a week and get back to permissions.

We have 109 rules created in 44 workspaces now.

Usability and simplicity

We seriously fucked up with Quick Filters where users can change values (especially me), as a result, we had to start from scratch and scrap 2-3 months of effort. However, this time we nailed a better solution with help of Anton and I hope Pavel will keep his sanity intact after this feature.

Another quite unexpected feature we’ve implemented is Architect vs. User modes. It is a somewhat controversial thing, and we are getting some negative feedback (but we got more positive feedback for sure).

What we learned here is to prepare releases for such features better. We had to add some in-product in-place announcement and make sure users are not missing this. An Intercom banner is not enough, so next time we will do better.

Done

We did many small improvements everywhere:

  • Quickly Show/Hide All Fields in Entity View
  • Reports performance boost: load less fields by default
  • Collapse and expand all levels for Grid View / List View / Gantt View
  • Hierarchical list: ability to remove top level objects with keeping the rest
  • Preserve all View settings when adding or removing databases where possible
  • View Items Menu unification
  • Group Action Buttons in context menu if there are more than 3 of them
  • Restyle collection XL units
  • Empty states for Views
  • Disable/Hide Documents for the workspace
  • Jump to newly created Field on Entity View
  • Update nested list styles in Smart Folders
  • Reorder action menu items in Entity page
  • Update Edit Name pop-up

Next

  • Timeline: improve performance to support lots of items
  • Search Spaces and other Search improvements
  • Hide Entity Views based on rules

Import and Integrations

We did several quite unexpected things

  • “Periods” integration
  • Stripe integration
  • Improve integration framework by reducing load on database

Next

Now we are working on some import improvements in scope of Introduce optional Field settings in integrations and import feature.

Files field

Victor was quite busy with different things and only now we are re-focusing on Files again. We also spent unreasonable amount of time on this bug File upload silently fails if you navigate away. File units got better:

  • New File unit on Entity View
  • Preview files in rich text fields in lightbox

Next

Now we will complete file units in other views (including Table View) and maybe take a look at Google Drive integration.

  • More File Unit visualizations on Entity View
  • File Unit visualization on Views
  • ??? maybe Attach Files from Google Drive to entities

Rich Text

We decided to make rich text great (for the first time), so Vova started to fix bugs and already fixed no less than 50 :lady_beetle:. Also Kostas did Embed Mermaid (code → diagram) into rich text.

Next

Kostas got some balls and will dig into Reorder rich-text blocks via drag’n’drop and keyboard. I hope we will be able to catch up with Notion here. We will continue bug fixing effort as well.

In parallel we are fixing references in Migrate mentions when converting or merging entity

AI

AI Agent usage increased significantly in January.

We’re spending $4K+/month on AI providers (mostly on Anthropic).

We can also calculate cost per account now, and in February we already see things are not going good for some accounts. We will think what we will do with AI pricing in some near future, since AI calls is not a good metric.

Sub-agents usage is also interesting. Guide is very popular.

Notable improvements that we’ve done recently:

  • Modify Document View content via AI
  • Search tool for AI Agent (incl. Document views)
  • Add Basic sub-agent for Multi-Agent system

Next

Aleh is completing Create/Update a Report with AI and now we are working on Set AI Agent Context (Space, Database, Entity, View, …).

Overall, Fibery AI Agent is far from complete and we have to implement many many improvements to make it good. We assembled a list of such things and Max with Sergey will move this theme forward in the next several months. We will also implement Smart Agent should access Files content (+ attach file into chat)

Automations

Very little progress here. We implemented Duplicate automations to other databases and nothing else.

Next

  • Access the previous value for Rules
  • Automatically link entities: operators besides = (>, <, ≠, …)
  • Automatically link entities: compare with constants
  • Unite “Entities linked/unlinked” triggers with Entity Update trigger, allow to mix all fields via OR condition

Whiteboard

To support a wider range of chart types in future Whiteboard: Templates , we started by expanding the set of connector styles and unifying the interface panels.
You can now use line jumps and a double-line style with all connectors, and we’ve updated the connector behavior algorithm to work better with different line thicknesses. Connectors also support custom colors and styles. When you select the connector tool, a new menu lets you choose the type, color, and style before you start drawing, so everything looks right from the start.

  • Redesign connector toolbar
  • Add connector a jump and a double line style
  • Update the set of arrowheads and line styles

Next

The next step is to unify the shape toolbars and add a sidebar for shape and icon collections, with support for user-defined collections.

  • Redesign shape toolbar
  • New panel for Shapes
  • Add the ArchiMate shape set
  • Add additional icon libraries

Mobile

By the end of December we shipped our first mobile app release to the app stores. In January we slowed down feature development to collect user feedback and focus on redesigning the mobile experience (navigation, document workflows, views, instant entity creation, and the AI chat and better tablet support).

P.S. Your feedback is welcome as usual!

10 Likes

This is high on my list - until we get Polymorphic relations, I have to frequently move entities out to another db because over time it turns out they belong elsewhere in a hierarchy. Since my team liberally uses comments (which generate references on other entities), and just references in general, it would be great if all that changed when converting was the actual location of the entity, and everything else in its history - in our case sometimes from a few years back - was preserved. Thanks for working on this!

Really looking forward to this update. I’d love to create a BUNCH of useful entity views for different use cases, but don’t want to overwhelm my average user with so many tabs on their entities! Being able to choose when a view is available will be really useful for my team.

PLEASE fix behavior when using numbered & bulleted lists in rich text fields – when deleting a bullet item that has items below it, the entire bulleted list below shifts indentation level & must be reset, line-by-line

It would be incredibly helpful (and I hope not too difficult) if the entity history panel included the name of rules that have modified it, instead of just “Fibery (automation rule)”. I’ve run into a couple instances where I can’t remember which rule applies to the entity in question & spend too much time tracking it down.

Overall very excited for the next few months of updates! Keep up the good work :slight_smile:

You said:

Fibery AI Agent is far from complete and we have to implement many many improvements to make it good

But I just want to say: it’s been a great feature for me, a new Fibery ‘architect’! You shouldn’t underestimate how valuable it already is - it saves me a ton of time trying to figure out things in docs, or puzzle through UI; I just ask the AI to do something for me, or ask it how a feature works, and that has reliably gotten me 90% of the way there. I just onboarded a bunch of new users and I encouraged them to use the AI if they weren’t sure how to do something.

6 Likes

Looking forward to these ones:

  • Hide Entity Views based on rules
  • Automatically link entities: operators besides = (>, <, ≠, …)
  • Automatically link entities: compare with constants
  • Whiteboard
  • Mobile
1 Like

Google Drive integration would be so wonderful! Between Airtable, SmartSuite and now Fibery, I’ve always wished this was a natively supported option. It seems crazy (for me) to pay for that central storage and not be able to easily use it in these systems. (Other than copying share links manually, which is what I’ve been doing.)

On the increased AI use, I think usage might go down with a review and cleanup of the docs. :slight_smile: I know this has been a topic before, so I won’t regurgitate it. But I would have been lost trying to get up to speed with API calls without the AI. If my experience as a newbie is not unique, the growing use of AI might be reflective of your growing number of new accounts. (A good thing!) Now that my system is basically up and functional and I have a handle on how it all works, I’m not using AI anymore. Maybe that’ll be the trend for others, too, thereby lowering your monthly costs. :thinking:

Finally, seeing “we seriously fucked up” in an official company post just makes me realize this is a good place to be. I seriously fuck up, too; it’s nice to see others admit it and make corrections!

I hope everyone has a shift-shift kind of day!

7 Likes

Sorry, we fixed this bug today 😛 February 5, 2026 / Pinned Filters where users can change values, Previous field values in validation rules