The June 2026 edition brings a wave of updates across the platform — headlined by Agent-to-Agent (A2A) support that lets external systems trigger your AI Colleagues automatically, unified Role-Based Access Control across AI Colleagues Studio, Workflows Studio, and Integrations, an AOP Migration Wizard for promoting processes from UAT to Production, and a new run-level AI Colleagues Analytics dashboard. Workflows Studio gains anytime AI editing and the ability to build list views and dynamic dropdowns from a prompt. Helpdesk introduces SLA Policies V2, a consolidated Helpdesk Overview, and self-serve notification preferences. Onboarding adds a candidate Checklist and a suite of Additional Settings, while platform updates bring a Group Chat Agent for Slack and a fully revamped Employee Sync experience.

In This Edition

---

AI Colleagues Studio

A major month for AI Colleagues — from letting external systems trigger your workflows automatically to unified access controls, safe promotion of processes across environments, and deep run-level analytics.

Agent-to-Agent (A2A) Support

Your AI Colleague can now work directly with other systems and agents — no human in the loop required. Until now, external apps connected to an AI Colleague through MCP, which works well when a person is asking the questions. But many real-world needs are system-to-system: a scheduler, an HRIS, or another AI agent that simply needs to kick off a Leena workflow on its own. That is exactly what A2A enables.

Admins turn it on with a single toggle in Orchestrator Settings, right next to MCP — credentials are shared, so there's no separate setup. Once enabled, external systems can trigger any business process directly and track its progress until completion. It's safe by design: access is locked to a single AI Colleague, and these runs stay fully separate from employee conversations. As a bonus, Leena can now plug into Microsoft Copilot Studio in two ways — as a tool employees use, or as a teammate agent that Copilot can delegate work to.

A simple way to think about it: if a person is asking, use MCP; if a system is triggering a known workflow, use A2A.

📖 Documentation: MCP Server + A2A · Copilot Studio Guide

Unified Role-Based Access Control (RBAC)

Access control is now unified across AI Colleagues Studio, Workflows Studio, and Integrations. Two new roles — AI Colleagues Studio Admin and AI Colleagues Studio Builder — can access all three areas from a single role. Every existing Workflows Studio Admin has been migrated to AI Colleagues Studio Admin, so they automatically gain access to AI Colleagues Studio and Integrations as well.

Alongside the access changes, a few things have been streamlined for clarity:

• Clearer terminology — The "Skills" section inside AI Colleagues Studio is now called Tools.
• Friendlier run options — For workflow-type tools, the execution options are renamed to Review in a form, Ask and run, and Ask, confirm and run.
• New scopes — Processes, workflows, tools, and connectors can now be Global (usable across all AI Colleagues) or AIC-scoped (usable only within their own AI Colleague). All pre-existing items default to Global, so nothing breaks.
• Register tools while building — You can now register a workflow as a tool directly while building it, with no need to jump between Workflows Studio and AI Colleagues Studio.

📖 Documentation: Role-Based Access Controls · Tool Configuration

AOP Migration Wizard — Promote from UAT to Production

You can now promote a business process from a test environment to production through a guided, two-phase wizard. Select a single process along with its linked tools and, optionally, its evaluation test cases; the destination AI Colleague is auto-matched for you, and a single mapping page walks you through every connection the process depends on.

The wizard is built to be safe and reliable. Pre-flight checks validate admin access, dependencies, and workflow availability upfront, clearly separating blockers that must be resolved from warnings that are simply surfaced. A failed migration saves nothing to production and can be retried cleanly, while a partial migration shows exactly what needs attention and re-runs only what failed. Migration is admin-only, with clear guidance for anyone who doesn't have the required access.

📖 Documentation: AOP Migration

AI Colleagues Analytics

A new run-level dashboard under Analytics gives you complete visibility into everything your AI Colleagues do. Track total runs and trends, along with participation coverage showing how much of your workforce is involved — from initiators to task assignees, escalation managers, and notification recipients.

Every run lands in a clear status (Handled, Unhandled, In Progress, or Cancelled), with escalation rates tracked separately and a reason attached to every unhandled run. Leaderboards surface the most active AI Colleagues, processes, and tools, and you can filter every panel at once by dimensions like status, escalation, trigger type, and more. Each panel drills down to a run-level view, and every row jumps straight into that run's step-by-step execution history.

📖 Documentation: AI Colleagues · Reporting · Run Execution

Note: AI Colleagues Analytics is currently released behind a feature flag and enabled per bot, with a broader default rollout planned soon.

---

Workflows Studio

The AI-powered workflow builder received a significant upgrade this month, making AI a full-time collaborator rather than a one-time generator.

Edit via AI — Anytime, Any App

AI editing is no longer limited to the initial creation session. Every editable app — including ones built manually — now has an Edit via AI button in the builder header that opens an AI chat panel right beside the canvas. You can mix manual and AI editing freely, even after publishing, and manual changes no longer disconnect an app from AI. The AI reads the current state of your workflow and adjusts it in place rather than rebuilding from scratch, and manual editing is only paused briefly while a change is being applied.

AI Builds List Views & Dynamic Dropdowns

The builder now understands more of what you're trying to create. It classifies your request into one of three app types — and tells you which one it picked:

• Custom App — form-based workflows with approvals and integrations.
• List View App — describe something like "show all pending requests in a table with an approve button per row," and the AI builds the columns, data flow, row actions, and filters for you.
• Dynamic Dropdown Utility — describe a dropdown that pulls values from a connected system, and the AI builds that data source end to end.

Expanded Node Support

AI generation now handles more advanced building blocks, including loops and asynchronous callback steps that can pause and resume a workflow. It can also look up external service details on the web while configuring an action, reducing the manual research needed to set things up.

📖 Documentation: Generate Workflows with AI

---

Helpdesk

A broad set of Helpdesk improvements this month spanning SLAs, agent productivity, notifications, and reporting.

SLA Policies V2

SLA Policies V2 brings a refreshed experience for configuring and managing SLAs, making policy administration simpler and more reliable. The redesigned interface streamlines how policies are created and managed, and the bulk upload experience has been significantly improved: new uploads now append to existing SLA configurations instead of replacing earlier data, validation errors can be reviewed before the upload is completed, and detailed error reporting points to the exact row, column, and issue so corrections are faster and easier.

📖 Documentation: SLA Policies V2

Note: SLA Policies V2 is being released behind a feature flag and will be enabled for all clients gradually, with the Product Team and CSMs coordinating the rollout.

Helpdesk Overview

Helpdesk Overview gives agents a single dashboard for everything on their plate — a consolidated view of all tickets accessible to them, in one place to see, analyze, and act. Tickets are organized into meaningful cohorts so agents can tackle the most urgent items first, such as tickets breaching SLA in the next two or four hours. The result is an at-a-glance picture of the day, less time spent hunting for what needs attention, and more efficient ticket resolution.

📖 Documentation: Helpdesk Overview

Self-Serve Notification Preferences

Notification management is now fully self-serve. Clients can configure notification triggers directly from the interface, with no backend dependency, and tailor exactly which scenarios send notifications for each role — Agent, Raised By, Raised For, and Watcher. Previously these preferences were managed on the backend; now teams can adjust notification behavior to fit their own workflows without raising a request.

📖 Documentation: Notification Preferences

Enhanced Reporting

Reports now carry more context for deeper analysis. Escalation details across L2 and L3 are included, giving better visibility into how tickets travel through escalation, and customer rating feedback comments are now available in reports, helping teams better understand user sentiment and service quality.

Refreshed Canned Responses

The Canned Responses page has a cleaner, more intuitive interface, making it easier to create, organize, and manage predefined responses.

Faster Report Downloads

Report generation and downloads have been optimized for better performance, especially when working with larger datasets.

Draft Comments During System Downtime

To prevent accidental data loss, comments entered by agents are now automatically saved as drafts when the system is unavailable, so work can be picked back up once connectivity is restored.

---

Onboarding

Two new capabilities give teams more control over the candidate onboarding experience.

Onboarding Checklist

Build the checklist that candidates use to track their onboarding tasks. Assemble a checklist from forms, notifications, and workflows, and drag items to set the order candidates see. You can set custom messages on form tasks for each submission status — Pending, Approved, Resubmit, or Not Submitted — and every checklist gets a shareable URL, along with enable, disable, and edit controls from the list.

📖 Documentation: Checklist

Additional Settings

A new group of advanced onboarding configuration pages is now available, all sharing a consistent view → edit → draft → publish flow. These cover candidate mapping, the hiring manager dashboard, dashboard configuration, stop reasons, email and reminders, document management, and more — each documented field by field.

📖 Documentation: Additional Settings

---

Platform & Global Settings

Updates to the channel experience and core employee data management.

Group Chat Agent for Slack

Your AI Colleague is no longer limited to one-on-one direct messages — it can now be mentioned directly in Slack channels, private channels, and group DMs, and it responds with full awareness of the ongoing conversation. Anyone in a channel can tag the AI Colleague with a question; it posts an instant acknowledgment and replies in the thread. Because it's context-aware, follow-ups like "summarize the issue above" simply work.

Delivery is smart about privacy: general answers are posted publicly in the channel with sources, while personal or multi-step answers are automatically redirected to the employee's direct message, so private data never lands in a public channel. The person asking always receives the complete answer in DM with a link back to the channel thread, and attachments shared with a mention are processed too.

📖 Documentation: Group Chat Agent for Slack

Note: Externally shared (Slack Connect) channels are not supported.

Revamped Employee Sync Experience

Configuring and managing employee sync in the Unified Dashboard is now much simpler. The old fragmented setup has been replaced with a clean, guided flow organized into four clear tabs — Manual sync, Automated sync, Sync history, and Employee data — with admins landing directly on Manual sync and mandatory fields pre-seeded to guide setup.

Automated sync has been rebuilt so each utility now has its own independent configuration with dedicated edit, activate, deactivate, and delete controls, and custom schedules that can run every set number of days, weeks, or months. New controls let admins choose whether a sync adds employees or only updates existing data, with built-in safety guards — including a block on any run that would terminate more than half of the synced population. Sync history is richer, showing added, updated, and terminated counts per run with filters and failure reasons, and the Employee data tab now offers active and terminated counts, column management, a full-record detail view, and one-click export.

📖 Documentation: Setup Employee Sync

---

Need help? Reach out to your Customer Success Manager or the Leena AI support team.