Release Notes

Every-two-week scheduling and a fresher Activity Log

This release adds an every-2-weeks option to Bloomy scheduled tasks, makes the Activity Log easier to keep current during an active device run, and polishes the login page. Bloomy automations can now follow an every-other-week cadence from a start week you choose, and the Activity Log gains a manual refresh, live updates while a run is in progress, and fresh data sooner.

What's new

Every 2 weeks for Bloomy tasks

Bloomy scheduled tasks can now run every 2 weeks, not just weekly, with control over which week the cadence begins.

• A new Every 2 weeks option sits alongside Weekly when you schedule a Bloomy task.
• A Starting date sets which week a biweekly task begins, so it lands on the right week from the first run.
• Task lists label these as "Every 2 weeks" so the cadence is clear at a glance.
• Ask Bloomy in chat to set up an every-other-week automation, and it schedules the matching cadence.

A fresher, steadier Activity Log

The Activity Log is easier to keep current during an active device run, and its loading behavior is more reliable.

• A refresh button in the Activity Log header pulls the latest results on demand.
• During an active device run, the Activity Log updates on its own so new events appear as they happen.
• Fresh activity now appears sooner, while a run is still in progress.
• Reset filters now clears Activity Log filters too, matching the other tabs.
• Fixed cases where changing filters during a refresh briefly showed outdated rows, or reverting them could leave the table loading.

Clearer login region switcher

The login page region switcher is clearer and easier to use.

• The region switcher shows proper US and EU flag images.
• Improved layout and larger tap targets make the switcher easier to use.
• A single, consistent loading indicator shows while you sign in.

Who benefits

• Teams automating with Bloomy: scheduled tasks that can run every 2 weeks from a start week you choose.
• Anyone scheduling in Bloomy chat: ask for an every-other-week automation and get the right cadence.
• People watching the Activity Log: on-demand refresh, live updates during a run, and fresh data sooner.
• Everyone signing in: a clearer login region switcher with proper US and EU flags.

Try it

• Open a Bloomy task and choose Every 2 weeks, then set a Starting week.
• Ask Bloomy to schedule an every-other-week report or automation.
• Use the refresh button in the Activity Log header, and watch it update during an active device run.
• Use Reset filters on the Activity Log tab to clear its filters.

Bloomy gets a memory, and delivers your reports

This release centers on Bloomy. It can now remember small, durable facts across conversations, list and deliver your saved Engineering Custom View reports with the same numbers you see on the dashboard, and run on the newest GPT-5.6 models. Scheduled Bloomy tasks also run more steadily, with fewer duplicate sends and clearer status when a run is skipped.

What's new

Bloomy remembers what matters

Bloomy can now hold onto small, durable facts you share, so you repeat yourself less from one conversation to the next.

• Bloomy remembers preferences and facts you tell it, and carries them across separate conversations.
• Review, and delete, everything Bloomy has remembered from My Bloomy.
• Scheduled tasks can carry their own remembered context, visible in the Task details dialog.

Bloomy delivers your engineering reports

Bloomy can now read your saved Engineering Custom View and executive reports and return the same numbers you see on the Custom Views tab.

• Ask Bloomy to list the saved Custom View and executive reports available to you.
• Ask Bloomy to summarize or deliver a named report for any date range, with numbers that match the Custom Views tab.
• Reports now include richer per-person breakdowns and clearer employee display names, with email shown only when you choose.
• Export a Custom View to an Excel workbook that includes every card, with consistent number formatting.
• Access follows the same permission as the Custom Views page.

Newer AI models in Bloomy

The Bloomy model picker adds the newest GPT-5.6 models and makes it easier to weigh intelligence against cost.

• GPT-5.6 Sol, Terra, and Luna are now available in the Bloomy model picker.
• Superseded models are retired from the picker and marked "(retired)" in existing chats.
• Model choices are grouped by provider, with clearer intelligence and cost indicators and fuller tooltips.
• Cost breakdowns for chats on retired models keep showing their historical spend.

Steadier scheduled tasks

Scheduled Bloomy tasks run more predictably, with fewer duplicate sends and clearer status.

• A scheduled task now knows what it already delivered, so repeat reports do not send the same thing twice.
• Tasks respect the data sources you currently have access to each time they run.
• When a scheduled task is skipped, you now see a notification instead of silence.
• Report attachments now arrive attached to the task email.

Who benefits

• Everyone using Bloomy: a Bloomy that remembers your preferences and facts across conversations.
• Engineering leaders: saved Custom View and executive reports delivered by Bloomy, with dashboard-accurate numbers and per-person detail.
• Admins choosing models: GPT-5.6 in the model picker and clearer intelligence and cost indicators.
• Teams running scheduled reports: fewer duplicate sends, visible skips, and attachments that arrive attached.

Try it

• Open My Bloomy to review and manage what Bloomy has remembered.
• Ask Bloomy to list your Engineering Custom View reports, then deliver one for a date range.
• Choose a GPT-5.6 model from the Bloomy model picker.
• Review a scheduled task's remembered context in the Task details dialog.

Sharper Usage Insights, from Activity Log to sprint delivery

This release centers on Usage Insights. The Activity Log gained richer filters, a fuller day timeline, and more accurate matching between devices and the people who use them. Engineering Insights adds sprint delivery metrics, admins can pull the current day's data on demand, and EU customers get clearer data residency defaults.

What's new

A richer Activity Log

The Activity Log in Usage Insights is easier to filter, read, and trust.

• Filter by device, application, and more with a collapsible filter row whose options build up as you page through results.
• A redesigned timeline shows a full view at once, with a summary row, an interactive legend, and times displayed in your own timezone.
• Turn on in-app title changes to expand any focus session into a per-title breakdown instead of separate zero-length rows, and export that same breakdown to CSV.
• Devices now map to the right person more reliably, which previously kept some devices out of team and person views.
• Selecting "Unassigned devices" now adds those devices on top of your current filters instead of clearing them, and operating-system and username filters behave correctly.

Sprint delivery in Engineering Insights

Engineering leaders can now measure how Jira sprints are landing, directly in Custom Views and executive reports.

• New sprint metrics cover completion rate, sprints completed, sprints in progress, sprint velocity, and ticket volume.
• A redesigned Sprints and Delivery tab on the Task Velocity card adds a toggle to include in-progress sprints, with plain-language tooltips.
• A sprint details view lists committed, completed, and carried-over issues, with board and date context.
• Metrics respect the selected time range, and completion rate counts only sprints finished within it, so story points match that same set of sprints.

On-demand real-time usage data

Admins can now have real-time access to usage data for their company when they need it, and scheduled Bloomy tasks report their status accurately.

• A Request Today's Usage Data action in Usage Insights runs the current day's data, shows progress, and delivers results to the History tab.
• On-demand runs now return in minutes rather than the longer wait they took before.

Bloomy reads your Activity Log

Bloomy can now answer questions grounded in your Activity Log, and longer conversations run to completion.

• Ask Bloomy to summarize activity logs, coverage, sessions, and more.
• Ask Bloomy to surface workflow patterns, context switching, and automation candidates from activity sessions for companies that have the optional opt-in window-title telemetry enabled.
• Access stays limited to companies and users with the Activity Logs module and the matching page permission.
• Longer, multi-step Bloomy conversations now run to completion instead of stopping early.
• Diagrams render more reliably, common formatting mistakes are repaired automatically, and view links open in a new tab.

Privacy controls and EU data residency

This release expands how teams control what is collected and where EU data is processed.

• Agent collection now has three clear states, collecting everything, collecting only what you allow, or collecting nothing, with explicit actions to turn collection on or off.
• Companies on the EU deployment start in a passive state, collecting nothing until an admin sets a policy.
• On the EU deployment, Bloomy defaults to EU-hosted models for data residency, with other providers available only when an admin opts in.
• Windows offers two installer choices, one with the browser extension and one without, and the setup page hides browser-extension controls for teams that do not use it.
• Windows device-agent updates no longer leave a duplicate entry in installed programs.

Who benefits

• Admins and leaders: a richer Activity Log with filters, a fuller timeline, and more accurate device-to-person matching.
• Engineering leaders: Jira sprint delivery metrics in Custom Views and executive reports.
• Admins: on-demand usage data, clearer collection controls, and two Windows installer choices.
• EU-based customers: passive collection by default and EU-hosted model processing for data residency.
• Everyone using Bloomy: activity-log answers and longer conversations that finish.

Try it

• Open Usage Insights, Activity Log and use the filters and redesigned timeline.
• Use Request Today's Usage Data in Usage Insights for a same-day pull.
• Review sprint delivery on the Task Velocity card in Engineering Insights.
• Ask Bloomy about your activity logs, and review collection controls in Integrations, Device Agents.

Simpler control over what device agents collect

This release refines the device-agent collection controls introduced last week. Setting the applications and websites an agent may collect is now more direct, the suggestions that help you build that list are available whenever you want them, and the setup page adapts to teams that do not use the browser extension. The changes live in Integrations, Device Agents.

What's new

Simpler collection controls

The controls for limiting what device agents collect are easier to set and easier to read at a glance.

• Filling an allowed list for an application or website card now turns collection limits on directly, and clearing the list collects everything again, with no separate step to manage.
• Each card shows its current state in plain terms, either collecting everything or collecting only the entries you allowed, so the limit in effect is never in doubt.
• A single Save changes control applies your edits and stays inactive until the list actually changes, with a confirmation that matches whether you are limiting collection or clearing it.

On-demand app and website suggestions

You no longer wait on a timed window to see what agents detect. Suggestions are now available on demand while you build a list.

• Pull a fresh list of the applications and websites agents detect from a refresh control on each card, whenever you need it.
• Suggestions list only the application and website names agents see, so you can build an allowed list without typing entries by hand or guessing.
• Applications stay governed separately for macOS and Windows, so each platform keeps its own list.

A cleaner setup page

The Device Agents setup page now reflects the collection choices a company has already made.

• Companies that turn off browser-extension collection no longer see the browser extension card during setup, and the desktop agent cards fill the row.
• The website collection card is hidden in that same case, since only the browser extension applies website limits.

Who benefits

• Admins configuring device agents: setting and clearing collection limits takes fewer steps, and each card states the limit in effect.
• IT and security leaders: allowed lists for applications and websites are easier to read and verify at a glance.
• Teams that do not use the browser extension: a setup page without cards and controls that do not apply to them.

Try it

• Open Integrations, Device Agents and use the Agents config tab to review your collection controls.
• Fill an application or website card's list to limit collection, or clear it to collect everything again.
• Use the refresh control on a card to pull the latest applications and websites agents detect.

Bloomy tasks that run when you expect them

This release is about scheduled work that behaves the way you set it up. When you edit a completed or archived Bloomy task, it now returns to your active schedule instead of quietly staying put. Teams in the EU region also get steadier document processing in Bloomy's knowledge base.

What's new

Bloomy scheduled tasks run after you edit them

Editing an archived or expired Bloomy task no longer leaves it stuck. Changes you save now put the task back on your active schedule.

• Editing a completed one-time task to run again on a future date now runs as scheduled, instead of staying marked complete and never firing.
• Saving edits to an archived task fully restores it to your schedule, instead of leaving it in a paused state that never runs.
• Restoring an archived task behaves the same whether you unarchive it or save your edits, so the result is consistent.

Steadier knowledge base for EU teams

Teams operating in the EU region see more reliable document processing in Bloomy's knowledge base.

• Documents added to the knowledge base now finish processing reliably for teams in the EU region.
• Re-indexing or updating existing knowledge base content completes as expected.

Who benefits

• Teams using Bloomy scheduled tasks: edits to completed or archived tasks return them to your active schedule.
• Admins managing recurring work in Bloomy: fewer tasks that silently stop running after an edit.
• Teams in the EU region using Bloomy's knowledge base: more reliable document processing when you add or refresh content.

Try it

• Open Bloomy and edit a completed task to a future date to see it return to your schedule.
• Restore an archived task from the task details view and confirm it runs on schedule.
• Add or refresh a document in Bloomy's knowledge base and watch it finish processing.

More say over what device agents collect

This release puts admins in control of what device agents gather and how they update. You can now define exactly which applications and websites agents are allowed to collect, give Windows fleets the same controls macOS already had, and manage all of it from a single Agents config tab. Sign-in gets cleaner too, with US and EU options so your team lands in the right place.

What's new

Decide what device agents collect

Admins can set the exact applications and websites their device agents are allowed to collect, and everything else is ignored on the device before it is ever stored.

• Choose which applications the desktop agent may collect and which websites the browser extension may collect, with anything off the list dropped at the source
• Run a time-boxed discovery period that lists only the app and website names agents see, with no usage, no user, and no device details attached, so you can build the list without having to guess or typing them by hand
• Move to enforcement when the list looks right, with a clear confirmation step and countdown timers you can extend or stop at any time
• Windows agents add an opt-in control for collecting window titles, off by default, and disclosed in the agent's own usage view when it is on
• Applications are governed separately for macOS and Windows, so each platform keeps its own list

Windows device agents reach macOS parity

The Windows device agent now supports the same admin controls that macOS already had.

• Pin or roll back the Windows agent to a specific signed version, matching the existing macOS control, with deployed agents picking up the target automatically and no reinstall
• Set menu-bar icon visibility and automatic updates separately for macOS and Windows
• Windows agents apply the pinned version more reliably, including a fix for updates that could get stuck part way through

All agent settings in one Agents config tab

Device agent controls now live together instead of being split across screens.

• A single Agents config tab holds the agent controls, including update behavior, menu-bar visibility, and version pin or rollback
• Every company with device agents sees these controls, not only those with the collection allowlist enabled
• The Overview screen is now a cleaner installer view with an at-a-glance row for employees, assigned agents, and agent status

A clearer sign-in with US and EU options

The sign-in page now points people to the correct environment and starts everyone on the same path.

• The login page shows US and EU options, so people reach the right environment instead of guessing web addresses
• Each company belongs to one region, which supports data residency for teams that need data kept in a specific region
• Account creation now begins with an email lookup, so you always land on the correct sign-in path for your company's enabled options

Who benefits

• IT and security leaders: decide exactly which applications and websites device agents may collect, with a discovery period before enforcement
• Teams with works-council or regional data rules: collection can be limited to an approved list, with everything else dropped on the device
• Admins running Windows fleets: pin or roll back the Windows agent and set per-platform visibility and update behavior
• Admins configuring agents: all agent settings now sit in one Agents config tab
• Everyone signing in: a US and EU switcher points you to the correct environment

Try it

• Open Integrations, Device Agents and use the Agents config tab to review agent controls in one place
• Turn on discovery for a set number of days to build your collection list, then switch to enforcement once it looks right
• Set the Windows agent version pin alongside the existing macOS control, and adjust per-platform menu-bar and update options
• Toggle window-title collection for Windows agents, off by default, if your policy allows it
• Use the US and EU switcher on the sign-in page to reach the correct environment