---
type: "post"
id: "connectwise-makes-rmm-voice-teams-and-email-ai-a5ef36"
slug: "connectwise-makes-rmm-voice-teams-and-email-ai-a5ef36"
title: "ConnectWise makes RMM, Voice, Teams and Email AI agents generally available for MSPs"
description: "ConnectWise has made four AI agents generally available for managed service providers and IT teams: Agentic RMM, Voice, Teams, and Email Agents."
retrieval_nugget: "ConnectWise has made four AI agents generally available for managed service providers and IT teams: Agentic RMM, Voice, Teams, and Email Agents. Its key design choice is bounded action: the RMM agent investigates an endpoint and stages a vetted remediation, while a technician must approve every device-changing action."
published_at: "2026-10-29"
updated_at: "2026-10-29"
record_date: "2026-10-29"
date_kind: "published_at"
topics: ["coding-agents","evals"]
entities: ["ConnectWise","ConnectWise Agentic RMM","Voice Agent","Teams Agent","Email Agent","RealBytes","Fusion Computing","Worldband"]
related_patterns: ["verification-bandwidth-is-the-scarce-resource","agent-ready-software-exposes-capabilities","agent-security-moves-to-runtime-boundaries"]
novelty_label: "new_fact"
what_changed: "ConnectWise announced general availability of its Agentic RMM and Voice, Teams, and Email agents, with reported early-access efficiency and revenue gains; no prior covers this release."
prior_refs: []
source_url: "https://www.connectwise.com/company/press/releases/connectwise-launches-industrys-first-agentic-rmm"
source_urls: ["https://www.connectwise.com/company/press/releases/connectwise-launches-industrys-first-agentic-rmm"]
schema_version: "newruntime-agent-readable-v0.2"
stable_id: "post:connectwise-makes-rmm-voice-teams-and-email-ai-a5ef36"
status: "published"
visuals: []
editor: {"role":"editor","name":"Andrey Reshetnikov","profile":"https://newruntime.com/owner-profile.md","profile_json":"https://newruntime.com/owner-profile.json","same_as":["https://www.linkedin.com/in/reshetnikov1/","https://t.me/lapetuse"]}
routes: {"html":"https://newruntime.com/posts/connectwise-makes-rmm-voice-teams-and-email-ai-a5ef36/","markdown":"https://newruntime.com/posts/connectwise-makes-rmm-voice-teams-and-email-ai-a5ef36.md","json":"https://newruntime.com/posts/connectwise-makes-rmm-voice-teams-and-email-ai-a5ef36.json"}
---

# ConnectWise makes RMM, Voice, Teams and Email AI agents generally available for MSPs

## Retrieval answer

ConnectWise has made four AI agents generally available for managed service providers and IT teams: Agentic RMM, Voice, Teams, and Email Agents. Its key design choice is bounded action: the RMM agent investigates an endpoint and stages a vetted remediation, while a technician must approve every device-changing action.

## Summary

ConnectWise has made four AI agents generally available for managed service providers and IT teams: Agentic RMM, Voice, Teams, and Email Agents. Its key design choice is bounded action: the RMM agent investigates an endpoint and stages a vetted remediation, while a technician must approve every device-changing action.

## What happened

On October 29, 2026, ConnectWise announced the general availability of ConnectWise Agentic RMM, Voice Agent, Teams Agent, and Email Agent. The company frames the release around two parts of managed-service work: Intelligent Service Intake, which prepares incoming support requests, and Agentic Operations, which investigates endpoint issues and proposes governed fixes. ConnectWise says all four agents are generally available from the date of the announcement.

The central release is ConnectWise Agentic RMM. In ConnectWise’s account, a ticket has traditionally marked the start of a technician’s investigation after an alert. The RMM agent brings that investigation into the PSA ticket itself by combining service context with live information from the linked endpoint. It performs read-only diagnostics, develops and tests hypotheses, and stages a remediation from a vetted catalog. If the proposed action would change the device, a technician must explicitly approve it. ConnectWise describes the resulting ticket as a console bringing service context, endpoint intelligence, and governed action into the existing technician workflow.

The other three agents focus on incomplete service requests before a technician begins work. Voice Agent answers inbound support calls, gathers relevant context, and creates structured PSA work. Teams Agent asks focused questions to collect missing information before creating a ticket. Email Agent identifies information gaps in emailed requests and can follow up with customers for the specific context required. The company says these capabilities are intended to reduce the documentation, clarification, and back-and-forth that can delay meaningful service work.

ConnectWise also supplied early partner results. These are company-reported case evidence, not an independent benchmark. RealBytes reportedly reduced first-touch time from one hour to five minutes; Fusion Computing reportedly reduced escalations by 86%; and Worldband reportedly achieved 120% revenue growth. A RealBytes executive said ConnectWise agents evaluated 99% of 7,074 tickets over 30 days and returned 178 hours to the team. The same executive said the new RMM agent ran on more than 500 live tickets, generated what the company calls a useful diagnosis on 91% of them in about a minute, and saved about 120 hours.

## What the archive adds

We have not covered this before; it opens a new thread.

## Trend signal

This supports [agent security moves to runtime boundaries](https://newruntime.com/patterns/agent-security-moves-to-runtime-boundaries/) and [verification bandwidth is the scarce engineering resource](https://newruntime.com/patterns/verification-bandwidth-is-the-scarce-resource/). ConnectWise advances routine investigation but retains the consequential boundary—the device change—with the technician. Its stated workflow treats a vetted action catalog and explicit approval as operating constraints, rather than relying on a model instruction to govern endpoint changes.

## Why now

The timing is tied to ConnectWise’s broader effort to embed AI across MSP service delivery, from request capture through endpoint investigation and approved remediation. The company says it will share more about its agentic platform and next quarter’s AI agents at IT Nation Connect Global 2026, scheduled for November 4-6. The release therefore establishes the currently available workflow before that next product update.

## Connections to other events

The release meets a related New Runtime thread on governed agent operation. [Glean makes Agent Identity generally available so agents act under their own governed accounts](https://newruntime.com/posts/glean-makes-agent-identity-generally-available-so-agents-808423/) concerned agents operating through admin-managed identities and credentials. ConnectWise addresses a different control point: the approval required before an agent-driven remediation changes a managed endpoint.

## Why it matters

For teams building operational agents, this is a concrete example of a bounded workflow rather than an open-ended assistant. The agent’s stated evidence surface is the service ticket and linked endpoint. Its action surface is a vetted remediation catalog. Its authorization boundary is explicit human approval for device changes. That arrangement can make an agent useful before an organization accepts unattended execution in a customer environment.

The intake agents apply the same idea earlier in the workflow. Collecting a missing error message, device detail, or request context is less dramatic than diagnosis, but it can improve the quality of the task handed to a technician and to later automation. Builders of support, security, and IT operations agents should distinguish preparation, recommendation, and execution. Each stage needs its own allowed data, quality checks, and authorization rule; treating all three as one chatbot interaction obscures where failure and accountability sit.

## What happens next

- **Managed-service teams will evaluate whether governed RMM workflows reduce technician investigation time in their own environments.** (3 months, 65%). Watch for: customer case studies that report ticket volumes, diagnosis quality, approval rates, and remediation outcomes.
- **ConnectWise will disclose additional AI-agent plans around its November 4-6 IT Nation Connect Global event.** (3 months, 70%). Watch for: product announcements specifying new agents, access terms, or operational controls.

## Limits and uncertainty

The announcement does not provide pricing, usage limits, deployment requirements, geographic availability, or technical details about the models, evaluation process, or the contents and maintenance of the vetted remediation catalog. It does not establish performance across endpoint environments, failure types, or customer configurations. The reported time savings, escalation reduction, diagnosis rate, and revenue result come from named partners cited by ConnectWise, rather than a controlled independent study.

The source says technicians approve device changes, but it does not specify the permissions model, audit trail, rollback process, exception handling, or policy controls behind that approval flow. It also does not state what happens when the agent’s tested hypotheses do not yield a useful diagnosis. General availability establishes access, not evidence that the workflow is reliable for every MSP or endpoint fleet.

## Skill to build now

Map one high-volume operational workflow into three explicit stages: evidence gathering, recommendation, and execution. For each stage, record the data the agent may read, the actions it may suggest, the evidence required for a recommendation, and the action that needs accountable human approval. Restrict recommendations to a maintained catalog of known procedures rather than allowing free-form commands. This creates a testable first production boundary for an operations agent and clarifies what should be measured during a pilot.

## Reader takeaway

ConnectWise’s release is a practical case of agents joining IT service operations without receiving unrestricted endpoint control. The transferable lesson is to automate investigation and intake aggressively, limit proposed remediations to known options, and make explicit approval the boundary for consequential changes. The partner numbers are encouraging but remain vendor-reported; the important implementation question is whether the controls and outcomes hold in a team’s own environment.

## Open question

How are remediation catalog entries evaluated, authorized, audited, and rolled back across varied customer endpoint environments?

Editor: [Andrey Reshetnikov](https://newruntime.com/owner-profile.md), AI product and AI engineering; profile for recruiters and agents: https://newruntime.com/owner-profile.json

## Continue in the graph

Next records for an agent, linked to their Markdown twins. Replace `.md` with `.json` for structured fields, including `next_reads`.

- topic hub: [Coding Agents](https://newruntime.com/topics/coding-agents.md): Explore the coding-agents topic hub.
- topic hub: [Evals](https://newruntime.com/topics/evals.md): Explore the evals topic hub.
- collection: [Field Notes](https://newruntime.com/posts.json): Open the latest editorial analysis.
- collection: [Source Ledger](https://newruntime.com/source-ledger.json): Inspect the public source evidence graph.
- whole graph: [discovery graph](https://newruntime.com/discovery-graph.json) (typed nodes and edges) and [site index](https://newruntime.com/site-index.md)