A Licensable Meeting Automation Layer for Workplace AI Platforms

Meeting automation looks simple from the outside.

A bot joins a meeting.
A transcript appears.
A summary gets generated.
Action items are extracted.
The meeting becomes searchable.

But underneath that experience is a difficult infrastructure problem.

To make meeting automation work reliably, a platform has to handle calendar logic, meeting-platform behavior, browser automation, bot orchestration, tenant isolation, transcript processing, AI summarization, searchable context, and cost control — all across the meeting tools modern teams already use.

With this release, xdge has rebuilt its Meetings infrastructure fully in-house.

The customer-facing experience remains the same. What has changed is the technology underneath: xdge no longer depends on a third-party meeting API layer. We now own the calendaring layer, meeting bot orchestration layer, and Playwright-powered browser automation stack needed to support meeting capture across major platforms.

This infrastructure is now available as a licensable technology layer for companies building workplace AI, enterprise search, productivity copilots, workflow automation, CRM intelligence, customer success platforms, and modern knowledge-management products.

Why Meeting Automation Is Harder Than It Looks

Most teams do not want to spend years rebuilding meeting automation from scratch.

The visible feature set is straightforward:

• Join a meeting
• Capture the conversation
• Generate a transcript
• Summarize what happened
• Extract action items
• Make the meeting searchable

The hidden infrastructure is much more complex.

A production-grade meeting automation layer has to solve for:

• Calendar discovery and scheduling logic
• Bot join timing and attendance behavior
• Browser automation across meeting platforms
• Platform-specific meeting-room behavior
• Authentication and tenant boundaries
• Capture reliability
• Transcript and summary processing
• Action-item extraction
• Searchable meeting context
• Cost-efficient scaling across many tenants

This is the infrastructure xdge has now built and standardized in-house.

What Changed

Before

xdge Meetings previously depended on a third-party meeting API vendor for key parts of the meeting automation stack.

That vendor layer helped abstract calendaring, meeting bot attendance, and meeting-platform connectivity, but it also introduced dependency, cost, and control constraints.

For any company trying to build meeting intelligence into its own platform, this is the typical tradeoff: either rely on an expensive third-party abstraction layer or build the entire meeting automation stack internally.

Now

xdge has replaced that third-party layer with its own infrastructure.

The Meetings stack is now powered by:

• In-house calendaring infrastructure
• In-house meeting bot orchestration
• Playwright-based browser automation
• Native control over meeting attendance flows
• Standardized tenant-level deployment
• Support for major meeting platforms
• Direct integration into downstream AI workflows such as transcripts, summaries, action items, and searchable meeting context

The end-user functionality stays the same. The underlying infrastructure is now owned and operated by xdge.

Built for Licensing, Embedding, and Platform Integration

This release is not just about xdge using better infrastructure internally.

It is about making meeting automation available as a licensable platform layer.

Companies building workplace AI products often need meeting intelligence, but do not want to build the plumbing themselves. That includes companies working on:

• Enterprise search
• AI copilots
• Workflow automation
• CRM intelligence
• Customer success platforms
• Sales intelligence
• Support intelligence
• Internal knowledge-management systems
• Productivity and collaboration platforms

For these companies, meeting data is one of the richest sources of enterprise context. But accessing it reliably requires far more than an AI summarization prompt. It requires the infrastructure to enter meetings, capture them, process them, and connect the resulting knowledge back into the product experience.

xdge now provides that infrastructure.

Instead of building calendar-driven bots, browser automation, meeting attendance, transcript processing, and searchable meeting context from scratch, partners can license the xdge Meetings infrastructure and integrate these capabilities into their own products.

Same Functionality, Stronger Foundation

This is a feature-parity infrastructure release.

The product experience remains consistent. The change is in ownership, control, and economics.

Functionality preserved

The Meetings platform continues to support:

• Calendar-based meeting capture
• On-demand meeting bot attendance
• Meeting bot joins across major platforms
• Meeting capture
• Transcript generation
• AI-generated summaries
• Action-item extraction
• Searchable meeting history
• Chat-style replay over meeting content

Infrastructure improved

The underlying platform now provides:

• Direct control over calendaring behavior
• Direct control over bot orchestration
• Browser-based attendance through Playwright automation
• Standardized rollout across tenants
• Reduced dependency on external vendors
• Lower cost to serve each tenant
• A stronger foundation for licensing and embedding

The experience remains familiar. The platform underneath is now more scalable, more controllable, and more cost-efficient.

Lower Cost Per Tenant

One of the most important outcomes of this release is cost structure.

Meeting automation can become expensive quickly. A single tenant may have many users, many calendars, and many meetings. When every meeting event depends on a third-party API vendor, the cost of serving that tenant scales through someone else’s infrastructure and pricing model.

By replacing the vendor layer with our own infrastructure, xdge removes the middle layer.

That dramatically lowers the cost per tenant and gives partners a more efficient path to embedding meeting automation inside their own products.

This matters for companies that want to build meeting intelligence into their platforms but need the economics to work at scale.

Standardized Across Tenants

This release also standardizes the Meetings infrastructure across tenants.

Instead of relying on a vendor-mediated meeting layer with external constraints, the platform now runs through a common xdge-controlled foundation.

Standardized capabilities include

• Calendar integration behavior
• Meeting discovery
• Bot scheduling
• On-demand bot joins
• Browser-based meeting attendance
• Meeting platform coverage
• Transcript processing
• Summary and action-item generation
• Searchable meeting context
• Tenant-level deployment and support

This makes the technology easier to operate, support, debug, package, and license.

A Building Block for Modern Workplace AI

Meetings are one of the highest-signal sources of workplace knowledge.

They contain customer pain, product decisions, sales context, support escalations, hiring feedback, project updates, roadmap tradeoffs, and executive alignment. But most of that knowledge disappears into recordings, scattered notes, or people’s memory.

A meeting automation layer turns that conversational knowledge into structured, usable enterprise context.

For companies building AI products around workplace intelligence, this is a critical capability. Meeting context can enrich search, copilots, workflows, CRMs, support platforms, customer success tools, and operational intelligence systems.

xdge Meetings infrastructure is designed to serve as that layer.

It captures the meeting, processes the knowledge, and makes the output available for downstream AI-powered product experiences.

Why This Matters

Building meeting automation in-house is expensive, slow, and operationally complex.

Relying entirely on a third-party API vendor can be faster initially, but it creates cost, control, and dependency constraints as usage scales.

xdge now offers a third path: a licensable meeting automation infrastructure layer that has already solved the hard parts.

Partners can use this technology to bring meeting intelligence into their own products without rebuilding the entire stack themselves.

Summary

With this release, xdge has rebuilt its Meetings infrastructure fully in-house.

The user-facing functionality remains the same, but the underlying platform has changed significantly. xdge now owns the calendaring layer, meeting bot orchestration layer, and Playwright-powered browser automation stack that powers meeting capture across major platforms.

This removes the third-party meeting API layer, dramatically lowers cost per tenant, standardizes behavior across tenants, and turns Meetings into a licensable infrastructure capability.

For companies building workplace AI, enterprise search, productivity copilots, CRM intelligence, customer success platforms, or knowledge-management products, xdge Meetings infrastructure offers a faster path to meeting automation without having to build it from scratch.

Overview
The platform is built to support complex, production-grade workflows across teams, systems, and AI-driven processes. This release focuses on strengthening that foundation, making workflows more consistent, outputs more reliable, and the overall system more stable under real usage. We’ve optimized how workflows execute, how reports are surfaced, and how integrations behave, so everything runs more predictably from build to execution.

⚙️ Workflow Execution: Predictable from Test → Production

Before

• Test runs didn’t always match production behavior
• Edits (especially multi-step changes) could apply inconsistently
• Draft and published workflows could drift out of sync

Now

• Test and live runs follow the same execution path
• Workflow state and configuration stay fully aligned
• Changes apply cleanly across both draft and published versions

Why it matters
You can iterate without second-guessing.
What you test is what actually runs.

📊 Reports: Accurate, Immediate, and Aligned

Before

• Reports sometimes lagged or required refresh
• Duplicate reports could appear during updates
• Reports didn’t always reflect the latest workflow state

Now

• Reports generate immediately after execution
• Each run produces exactly one report
• Reports stay tied to the exact version of the workflow that ran

Why it matters
No more guessing what happened, reports now reflect reality, instantly.

🔌 Integrations: Reliable Across Systems

Before

• Slack messages could fail or format incorrectly
• Gmail workflows were inconsistent across runs
• Timezone/date handling caused incorrect outputs

Now

• Slack delivery and formatting are consistent
• Gmail-based workflows execute reliably end-to-end
• Time and date logic is standardized across integrations

Why it matters
Workflows now behave reliably outside the platform, not just inside it.

🤖 AI Workflows: From Variable → Repeatable

Before

• Deep Research outputs varied significantly between runs
• References and results didn’t always load consistently
• Prompts didn’t reliably map to outputs

Now

• Repeated runs produce more consistent outputs
• References and results render reliably
• Prompt-to-output alignment is tighter and more predictable

Why it matters
AI workflows are now stable enough for real use cases — not just experimentation.

⚡ Performance: Faster Under Real Load

We’ve grouped a set of performance improvements that collectively impact day-to-day usage:

• Faster workflow execution times
• Improved responsiveness under concurrent usage
• Reduced interruptions in integration-heavy workflows

Why it matters
The system holds up under actual usage, not just isolated runs.

🧩 UI & Interaction: Less Friction While Building

We’ve refined core interaction patterns across the platform:

• Clearer naming across workflows and components
• More consistent behavior across create, edit, and draft flows
• Improved UI responsiveness and state accuracy

Why it matters
Building workflows feels more predictable and less error-prone.

🛠️ Stability Fixes

We’ve resolved a set of edge cases affecting execution, reporting, and UI behavior:

• Fixed inconsistencies in workflow execution and report generation
• Improved tool visibility and behavior across workflows
• Resolved UI state issues across editing and chat flows

Why it matters
Fewer edge cases → more trust in daily usage.

Overview

The platform has always been designed to support complex, multi-step workflows across teams and systems. This release focuses on refining that foundation—improving execution consistency, strengthening system reliability, and making outputs more dependable across every run.

We’ve tightened how workflows execute, how reports are persisted and accessed, and how integrations behave under load. The result is a system that feels faster, more stable, and more predictable—whether you’re iterating in development or running workflows at scale in production.

🛠️ Workflow Execution & Consistency

This release improves how workflows execute across edits, reruns, and concurrent usage.

• More consistent execution across repeated runs and workflow updates
• Improved handling of long-running and parallel workflows
• Reduced variability when modifying workflows mid-iteration
• More stable behavior under concurrent load

Result: workflows maintain consistent behavior across iterations and scale more reliably as usage increases.

📊 Reports & Workflow Alignment

We’ve refined how reports are generated, stored, and surfaced so they stay aligned with workflow execution.

• Test and production reports are now consistently available in the Reports tab
• Draft and published states are clearly maintained
• Reports appear reliably without requiring refresh
• Improved visibility into outputs generated during testing

Result: reports remain accessible and aligned with workflow state, making it easier to review outputs and validate behavior across iterations.

🔌 Integration Improvements

This release strengthens how integrations perform across supported connectors.

• More consistent Slack message delivery and formatting
• Improved responsiveness across integrations
• More reliable data retrieval from connected systems
• Reduced latency across integration calls

Result: integrations behave more consistently and respond more reliably across workflows.

🤖 AI Workflow Behavior

We’ve improved how AI-driven steps behave across workflow edits and reruns.

• Better retention of workflow context across iterations
• More predictable outputs across similar inputs
• Improved consistency when modifying workflows

Result: AI-driven workflows produce more stable and repeatable outputs over time.

⚡ Performance Improvements

Performance has been optimized across workflows and integrations.

• Faster execution for long-running workflows
• Smoother performance across concurrent runs
• Reduced latency in integration-heavy workflows

Result: workflows execute more efficiently and scale more smoothly under load.

🔧 System Stability

We’ve made underlying improvements to system coordination and reliability.

• Improved consistency across workflow and report handling
• More stable behavior across system processes
• Stronger performance at scale

Result: the platform operates with greater stability and consistency across environments.

Overview

This release is a step-function improvement in how workflows run, how systems respond, and how reliable the platform feels end-to-end.

We’ve eliminated sources of inconsistency across execution, reporting, and integrations—tightening the system so workflows behave as expected, reports stay in sync, and connectors deliver with precision. The result is a platform that’s not just functional, but dependable under real usage.

🛠️ Major Improvements

Workflow Reliability & Execution

Workflows now execute with significantly greater stability across edits, reruns, and concurrent usage.

• Deterministic behavior when running and updating workflows
• Stronger handling of long-running and parallel executions
• Reduced failure points and unexpected interruptions

Reports & Draft Experience

Reports now reflect workflow state with accuracy and consistency—without manual intervention.

• Clear and reliable representation of draft vs. published states
• Reports persist correctly without refresh or loss
• Improved visibility into outputs generated during testing

Integration Improvements

Integrations now operate with higher reliability and tighter response cycles.

• Consistent Slack delivery and structured message formatting
• Faster, more responsive integration performance
• Increased accuracy across connector-based data flows

AI Workflow Behavior

Workflow logic is now more stable and context-aware across iterations.

• Predictable behavior when editing or rerunning workflows
• Stronger retention of prior workflow context
• More consistent and repeatable outputs across runs

UI & Experience

The interface has been refined to reduce friction and improve clarity across key workflows.

• More cohesive layouts and visual consistency across pages
• Improved spacing, alignment, and component structure
• Smoother, more intuitive interactions across the platform

⚡ Performance Enhancements

Workflow & System Performance

Performance improvements ensure workflows scale with usage without degradation.

• Optimized execution for long-running workflows
• Faster response times across integrations
• Stable performance across concurrent workflow activity

🔧 Infrastructure & Operations

System Stability Improvements

Underlying system upgrades reinforce reliability across all platform layers.

• More consistent handling of workflows and reports
• Improved coordination across internal system processes
• Stronger performance under sustained and scaled usage

Overview

This release strengthens workflow reliability, improves connector stability, and enhances report visibility across the platform. We focused on ensuring workflows run consistently during long executions, improving accuracy across Slack and Microsoft integrations, and making reports easier to access and track. These updates make automation more dependable, connectors more responsive, and overall workflow performance more stable.

🛠️ Major Improvements

Workflow Intelligence

Enhanced Workflow Search Capabilities
Workflows now support a hybrid of keyword and semantic search, enabling more precise information retrieval across your workspace.
What’s improved:

• Combined keyword + semantic retrieval
• More accurate discovery across connectors
• Stronger context grounding for workflow execution

Smarter Tool Orchestration
The workflow engine now prioritizes existing tools and integrations instead of generating redundant API logic.
Updates include:

• AI agents detect available tools before generating new code
• Improved orchestration across connectors
• Reduced unnecessary code generation
• More reliable and maintainable automation behavior

Platform Improvements

Workflow Execution Stability
Enhancements improve reliability when running multiple or long-running workflows.

• Improved handling of extended workflow executions
• Workflows no longer disappear after browser refresh
• More stable execution when running workflows simultaneously
• Consistent workflow progress tracking

Reports & Drafts Experience
The reporting experience has been refined to ensure outputs are always visible and accessible.

• Draft reports consistently appear in the Reports tab
• Notification badges now indicate new draft reports
• Reports generated during workflow tests are now visible
• Eliminated cases where draft reports disappeared during repeated executions

Integration Improvements

Slack Integration Stability
Slack interactions are now faster and more reliable.

• Improved handling of Slack API rate limits
• Reduced delays in typing indicators
• More reliable message retrieval and interaction

Microsoft Connector Improvements
Fixes and enhancements across Microsoft integrations improve accuracy and consistency.

• Improved Outlook email search accuracy
• Resolved email categorization issues
• Fixed Microsoft Teams notification inconsistencies
• Enhanced OAuth token handling

🐞 Bug Fixes

Workflow & Automation

• Fixed duplicate workflow creation issues
• Resolved workflows disappearing after extended execution
• Fixed unexpected steps when editing workflow time ranges
• Addressed failures in workflows shared across workspaces
• Fixed Gmail search failures during workflow verification
• Resolved Google Calendar workflow test failures

AI Workflow Reasoning

• Fixed incorrect detection of unavailable Jira and Slack tools
• Removed unwanted Jira-related instructions in generated release notes

UI & Experience

• Fixed Drafts page loading placeholders getting stuck
• Corrected notification badge display issues
• Resolved profile section inconsistencies
• Fixed layout issues on smaller screens (13-inch displays)
• Corrected tooltip truncation for long prompts
• Resolved homepage layout overlap when launching prompts
• Updated connector labels for consistency across the platform

⚡ Performance Enhancements

Workflow & Connector Performance

• Improved reliability across long-running workflows
• Faster and more stable Slack and Microsoft interactions
• Improved responsiveness in workflow execution and tracking

🔧 Infrastructure & Operations

System Reliability Improvements

• Strengthened workflow execution handling across edge cases
• Improved internal handling of connectors and authentication flows
• Increased stability for multi-workflow and long-running operations

Extending xdge Skills

When we first launched Skills by xdge, it consisted of two core building blocks:

• Workflows — structured, multi-step automation
• Prompts — reusable reasoning units for search, analysis, and generation

Both were powerful. Both were flexible.

But neither had direct access to a stable, production-grade action layer. If a Workflow or Prompt needed to take action, post to Slack, update Salesforce, or create a document, the system dynamically generated the required tool at runtime.

It worked. It felt magical. But it wasn’t how production systems scale.

Today, we’re introducing Tools, a third dimension of Skills, launching with 100+ prebuilt, certified actions that Workflows and Prompts can now call directly.

The Original Model: AI Generating Tools at Runtime

In the first generation of Skills, execution worked like this:

You describe what needs to happen. The system reasons through it.
If a required tool didn’t exist, the code generator would build it on the fly.

Under the hood, it would:

• Interpret the API
• Infer input/output schemas
• Generate integration logic
• Test execution
• Retry if needed
• Run safely

This unlocked enormous flexibility. We weren’t constrained by a fixed catalog of actions, the system could synthesize what didn’t exist.

But there was a structural cost.

Every time a Workflow ran, for every tenant and every user, parts of the execution layer had to be rebuilt in real time. Even with guardrails and validation loops, that introduced latency and variability.

Common workflows could take 10–15 minutes to complete.
Results were correct, but execution wasn’t deterministic.

For enterprise systems, that distinction matters.

Introducing Tools as a Third Dimension of Skills

Until now, Skills had two dimensions:

• Workflows — orchestration
• Prompts — reasoning

That combination allowed teams to analyze and coordinate work across systems.

What it lacked was a persistent, production-grade execution layer.

Tools are that third dimension.

Instead of dynamically generating common tools at runtime, we:

• Took the same reasoning engine used in dynamic generation
• Ran it offline at scale
• Put developers in the loop
• Standardized schemas and I/O contracts
• Hardened and QA-certified behavior
• Implemented directly in our backend

The result: 100+ production-grade, prebuilt tools available out of the box.

Most importantly:

Workflows and Prompts can now call Tools directly.

Reasoning lives in Prompts.
Orchestration lives in Workflows.
Deterministic execution lives in Tools.

Architectural Shift: Prebuilt Core Actions + AI for Custom Logic

Here’s what changed structurally.

Previously

• Skills = Workflows + Prompts
• Actions generated dynamically at runtime
• Tool synthesis embedded inside execution loops
• Retry and validation inside code-generation cycles

Now

• Skills = Workflows + Prompts + Tools
• High-frequency actions implemented as backend services
• Structured and standardized input/output contracts
• Stable, repeatable execution paths
• Code generator focused on long-tail and custom logic

Repeatable, high-frequency work is handled by certified backend Tools.

The AI generator is reserved for:

• Custom glue logic between systems
• Rare edge cases
• Specialized transformations
• Novel workflows that don’t justify a prebuilt primitive

This separation is intentional. It’s how you get speed and flexibility without sacrificing reliability.

10–15 Minutes to 1–2 Minutes: Faster, Predictable Execution

With Tools in place:

• Workflow execution drops from 10–15 minutes to 1–2 minutes for common paths
• Results are consistent across tenants and runs
• Observability improves
• Enterprise trust increases

Workflows feel faster because they no longer rebuild their action layer on each run.

Prompts become more powerful because they can now trigger deterministic execution directly, not just reason about outcomes.

AI moves from analysis to action without introducing variability into the system.

What’s Included in the 100+ Tool Launch

The first 100+ Tools focus on high-frequency, business-critical actions across the systems teams use every day.

Salesforce

• Change opportunity stage
• Update opportunity fields
• Create or update tasks
• Add notes or follow-ups

Jira

• Add a comment
• Change issue status
• Reassign an issue
• Update fields

Gmail

• Add or remove labels
• Draft an email
• Send templated replies
• Organize threads

Slack & Microsoft Teams

• Post a message to a channel
• Send direct messages
• Trigger structured updates

Documents & Content

• Create Google Docs
• Generate Word documents
• Structure documents from templates
• Populate content dynamically

These are the actions that typically follow reasoning:

Analyze → Decide → Act

Now, those action steps are prebuilt, certified, and directly callable.

Browsable in skills.xdge.ai

Executable in app.xdge.ai

Tools aren’t hidden infrastructure. They are visible, discoverable building blocks inside Skills.

You can:

• Browse the growing catalog at skills.xdge.ai
• Understand each tool’s input/output contract
• See examples of usage
• Call them directly inside app.xdge.ai via Workflows or Prompts

This makes Tools durable assets, not ephemeral code generated during execution.

They are searchable.
They are versioned.
They are reusable across teams.

And because they’re standardized, every Workflow and Prompt can rely on deterministic behavior.

From AI to Infrastructure

The first version of Skills proved that natural language could orchestrate real work across enterprise systems.

This version makes it dependable.

AI can generate impressive results.
Infrastructure delivers consistent outcomes.

By introducing Tools as a certified execution layer inside Skills, we’ve shifted from dynamic generation everywhere to a structured, layered system:

• Prompts handle reasoning
• Workflows handle orchestration
• Tools handle deterministic execution
• Common actions are prebuilt and production-grade
• Custom logic is generated only where flexibility is required

This is how AI platforms mature, not by removing flexibility, but by putting structure around it.

Scaling the Tools Layer

Today: 100+ prebuilt tools.
Next: hundreds more.

We’re expanding across connectors and deepening capabilities inside each integration. The roadmap moves from dozens of apps toward 100+ apps and 1,000+ connectors.

The vision remains the same:
Natural language as the control plane for enterprise systems.

What changed is the execution foundation.

If you’ve ever thought, “This should be automated.”

You’re right.

Now, Workflows and Prompts can call certified Tools directly and execute fast enough, and predictably enough, to depend on them.

Overview

This release focuses on reliability, automation stability, and meaningful UX refinements across xdge. We’ve resolved key issues affecting scheduled workflows, chat behavior, meeting status, and notifications while improving overall consistency across search, navigation, and editing experiences.

The result: workflows run when they’re supposed to, chat behaves predictably, and the product feels faster and more polished across the board.

✨ Improved User Experience

More Reliable Scheduled Workflows

Automation is now more dependable:

• Scheduled workflows execute at the exact configured time
• Scheduled runs no longer fail silently
• Workflow status updates correctly after execution

You can trust your workflows to run without manual intervention.

Smoother Workflow Editing

Workflow editing is now clearer and more stable:

• Renaming a workflow no longer causes unexpected regeneration
• Workflow steps display clearer, human-readable titles
• Editing instructions now updates correctly across views

Creating and refining workflows now feels more intuitive and predictable.

Cleaner Search & Navigation

We’ve improved search behavior and UI responsiveness:

• Long search queries render correctly
• Search animations and tooltip delays are faster
• Pagination improvements for workflow tiles
• Improved spacing and layout consistency across pages

The interface now feels more responsive and cohesive across screen sizes.

Meetings & Recording Improvements

Meeting reliability and clarity have been improved:

• Fixed incorrect “recording in progress” indicators
• Improved meeting recording timing reliability
• Optimized empty-state layout for smaller screens

Meeting status now reflects real activity accurately.

🐞 Bug Fixes

Chat & Notifications

• Fixed issue where follow-up responses could overwrite earlier messages
• Resolved chat scroll behavior so new questions anchor correctly
• Fixed notification unread indicator alignment
• Corrected small-screen layout overlap issues

Chat and notifications now behave consistently and render cleanly.

Workflow Stability

• Fixed issues that could impact scheduled execution timing
• Resolved workflow notification inconsistencies
• Improved tooltip clarity and edit button visibility
• Standardized pop-up and filter styling across views

Overall stability and visual consistency have been strengthened throughout the platform.

Overview

This release (Jan 15–31) focuses on strengthening workflow reliability, improving report accuracy, and refining the overall user experience across the platform. We’ve addressed key issues affecting workflow execution, report consistency, and integrations, while introducing improvements to notifications, Clips, and UI responsiveness. As a result, workflows run more reliably, reports are more accurate and complete, and the platform feels faster and more intuitive across all touchpoints.

🔔 New Features

Enhanced Notification System

Stay on top of workspace activity with a faster, clearer notification experience.

• Real-time system and workspace notifications via the bell icon
• Improved formatting for better readability and prioritization
• More accurate delivery of important updates

Trial Expiration Email Alerts

Workspace admins now receive clear email notifications when trials expire.

• Ensures visibility into subscription status
• Provides clear next steps for renewal

🛠️ Major Improvements

Workflow Reliability & Execution

Workflows now execute more consistently, with improved handling of edge cases and integrations.

• Fixed failures in Slack-based workflows and connector-driven steps
• Improved handling of workflow execution states and transitions
• Enhanced stability for workflows using multiple integrations (Jira, Google Docs, Salesforce)
• Resolved issues with stale or inconsistent workflow artifacts
• Improved workflow recovery and retry handling for failed executions

Reports & Data Accuracy

Reports are now more reliable, complete, and aligned with workflow outputs.

• Fixed issues where reports required refresh to appear
• Improved inclusion of data sources (e.g., Google Calendar events now reliably appear)
• Resolved inconsistencies in report sections across integrations
• Improved visibility and consistency of workflow-generated reports

Workflow Search & Integrations

Search and connector behavior is now more accurate and dependable.

• Improved Slack activity and mentions retrieval in workflows
• Fixed cases where connected apps were not properly recognized
• Improved reliability of connector-based search results
• Reduced failures across Jira, Gmail, and Salesforce integrations

Clips & Meetings Experience

Clips and meeting workflows are now more reliable end-to-end.

• Clips now open automatically in a new tab after recording
• Improved clip processing and playback reliability
• Fixed issues with missing or corrupted recordings
• Improved meeting bot reliability and transcript capture

UI & Experience

We’ve refined the interface to improve clarity, responsiveness, and consistency.

• Improved layout and responsiveness across screen sizes
• Fixed tooltip behavior, dropdown inconsistencies, and UI alignment issues
• Improved chat responsiveness and reduced UI blocking issues
• Standardized button styles, labels, and overall visual consistency
• Reduced visual clutter and improved readability across pages

🐞 Bug Fixes

Workflow & Automation

• Fixed workflow reports not appearing without refresh
• Resolved app-switching issues that could break workflows
• Fixed workflow plan getting stuck during generation in specific cases
• Improved workflow debugging and execution visibility

Reports & Outputs

• Fixed missing data in reports (including calendar events)
• Resolved incorrect report sections appearing across integrations
• Improved consistency when copying outputs (e.g., citation formatting fixes)

Integrations

• Fixed Jira reconnect issues and connector inconsistencies
• Resolved Slack workflow failures and search issues
• Improved Gmail and Salesforce workflow reliability

UI & System Behavior

• Fixed left-side chat UI getting stuck in certain scenarios
• Resolved dropdown duplication and layout inconsistencies
• Fixed icon, label, and alignment issues across the platform
• Addressed onboarding and connector visibility issues

⚡ Performance Enhancements

Workflow & System Performance

The platform now performs more reliably across workflows, search, and integrations.

• Faster workflow execution and report generation
• Improved responsiveness across workflow builder and reports
• More stable performance across concurrent workflows

🔧 Infrastructure & Operations

System Stability Improvements

Underlying improvements ensure more consistent performance at scale.

• Enhanced handling of workflow state transitions and retries
• Improved reliability of connector interactions and data retrieval
• Strengthened overall system stability across workflows, reports, and UI

🧹 Deprecations & Removals

Naming Update

• “Deep Research” has been renamed to Workflows
• Existing links continue to redirect, but updating bookmarks is recommended

Trial Changes

• Free trials have been removed for new signups
• Existing customers remain unaffected