AI Copilot

Helm pairs every technician with an AI copilot that understands Windows, PowerShell, and MSP workflows. You stay in control — the AI suggests, you approve.

How It Works

Describe the problem in plain English. The copilot translates your intent into the right diagnostic or remediation commands, displays them for your approval, and explains the results when they come back.

Example prompts:

“Check why QuickBooks keeps crashing”
“Find out what’s eating disk space on this machine”
“Why is this user’s Outlook so slow?”
“Something is killing the network performance — find it”

Human-in-the-Loop by Default

Every command the copilot generates is shown to you before execution. You review, approve, modify, or reject. When a command fails, the copilot explains the error and suggests a corrective action — you decide whether to proceed.

AI-only mode (where the copilot acts autonomously) is on our roadmap as a future toggle. Today, the human is always in the loop.

Full Context, Zero Ramp-Up

The copilot operates with complete context from your RMM and PSA, visible inline during every session:

Device specs, OS version, installed software (from your RMM)
Open PSA ticket details — title, description, prior notes
Recent RMM alerts: disk, CPU, memory, event log flags
Microsoft 365 account status (when M365 integration is enabled)

Your copilot already knows what is happening on the endpoint before you type a word.

Active Error Explanation

When a command returns an unexpected result or fails outright, the copilot does not just show you the error. It explains what went wrong, why it likely happened, and what to try next — all in plain language.

Supported Models

Helm uses a BYOK (Bring Your Own Key) model for Pro and Enterprise tiers. You supply your own LLM API key via OpenRouter, Anthropic, OpenAI, or any OpenAI-compatible provider. The Free tier includes a budget model at no cost to you.