AI Troubleshooting
AI Troubleshooting Assistant
Section titled “AI Troubleshooting Assistant”Helm’s AI troubleshooting engine understands Windows, PowerShell, and MSP workflows at a Tier 3 level.
How It Works
Section titled “How It Works”Describe the problem in plain English. The AI generates the right diagnostic commands, runs them against the endpoint, interprets the output, and proposes next steps — automatically.
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”
Command Execution
Section titled “Command Execution”Every command the AI generates is shown to you before it runs (or in auto mode, executed immediately). Output is grouped into visual blocks — you can see exactly what ran and what came back.
When a command fails, the AI explains why and suggests the corrective action.
Context Awareness
Section titled “Context Awareness”The AI operates with full context:
- Device specs, OS version, installed software
- Open PSA ticket (title, description, prior notes)
- Recent RMM alerts (disk, CPU, memory, events)
- User’s Microsoft 365 account (if M365 integration is enabled)
This means it doesn’t start from zero. It already knows what’s wrong before you type a word.
Supported Models
Section titled “Supported Models”Helm uses a BYOK (Bring Your Own Key) model for Pro and Enterprise. You supply your own LLM API key via OpenRouter, Anthropic, or OpenAI. Recommended: Claude Sonnet or GPT-4o via OpenRouter.