How to Train Atera’s AI Copilot for Personalized Support Responses

Atera’s AI Copilot delivers contextual ticket suggestions and automated resolutions, but its true power lies in tailoring those responses to your unique environment. By feeding Copilot the right inputs—knowledge base articles, scripts, contacts, and domain data—you ensure it offers personalized, accurate support. This guide shows IT managers and MSP leads exactly how to “train” Copilot using built-in tools and insights, without writing a single line of AI-specific code.

Key Summary

• Use the AI Center’s Actionable Insights to expand Copilot’s knowledge with your KBAs and scripts 
• Sync contacts and domains for up-to-date end-user context, boosting personalized replies 
• Assign agents and phone numbers to end users to enable accurate troubleshooting actions 
• Provide feedback via the Help menu (live chat, idea submissions) to refine AI behavior 
• Monitor Optimization Status in AI Center to track completion and identify gaps  

1. Leverage Actionable Insights to Expand AI Knowledge

Copilot bases its responses on the resources it “knows.” The Actionable Insights tab in AI Center surfaces suggestions for new scripts and knowledge base articles (KBAs) relevant to recurring ticket trends.

1. Navigate to AI Center > Actionable Insights.
2. Under Available, expand each insight to preview the suggested KBA or script.
3. Click Review article or Review script to inspect content.
4. Implement the insight by adding it to your KB library or script catalog.

Why it matters: Every implemented KBA teaches Copilot how to answer similar questions, while each script added under AI actions gives Copilot real remediation steps it can recommend or execute. 
 

2. Enhance AI Actions by Assigning Agents & Syncing Contacts

Personalized support requires Copilot to recognize users and assign context-aware tasks. Under AI Center > Optimize, focus on the AI Actions section:

• Assign agents to end users: Ensure each contact in your PSA has an assigned technician. Copilot then suggests or runs actions under the correct user context.
• Add phone numbers: SMS-based password resets and notifications rely on valid mobile numbers.
• Sync with Azure AD & Microsoft 365: Up-to-date directory data prevents stale or missing user records.
• Connect domains: For on-prem password resets and domain-bound tasks, link your Active Directory domains.

Tip: Walk through each AI Actions sub-tab and aim for at least 80 % completion to maximize Copilot’s ability to suggest and perform device actions. 

3. Provide Continuous Feedback via the Help Menu

Beyond automated insights, Copilot adapts when you share feedback directly:

• Click the Help icon on any Atera page (bottom right).
• Use Live Chat to report unexpected AI suggestions or request feature enhancements.
• Submit ideas and feedback—each entry trains the underlying models.

Note: These channels feed into Atera’s product team and AI training pipeline, influencing future Copilot updates. 

4. Monitor Your Optimization Status

The Optimization Status gauge in AI Center provides a real-time view of your Copilot setup:

• Site completion shows overall progress for your account.
• Customer completion tracks per-client readiness (ideal for MSPs).

Review the breakdown to see which recommendations—KBAs, scripts, agent assignments, or integrations—are still pending. Prioritize the highest-impact items to drive rapid gains in response accuracy and automation coverage. 
 

5. Job-to-Be-Done Scenario: Tailoring Copilot for a Growing MSP

Imagine you’re the MSP lead supporting 200 devices across five mid-market clients. Each client uses custom ticket categories and has distinct escalation paths. By:

1. Implementing client-specific KBAs suggested in Actionable Insights,
2. Assigning dedicated agents per client in the Optimize > AI Actions tab,
3. Syncing each client’s Azure AD,
    

you teach Copilot to:

• Suggest the correct category for each ticket,
• Provide client-accurate instructions,
• Trigger scripts that respect individual maintenance windows.

The result? A Copilot that “speaks your client’s language” and lets you scale support without extra headcount.

Ready to Refine Your Copilot?

Click Here

FAQ

How often should I review Actionable Insights in Atera AI Copilot?

You should review Actionable Insights weekly to catch new patterns in recurring tickets and keep Copilot aligned with emerging issues. This helps ensure Copilot suggestions stay relevant and tailored to your environment.

Can Atera Copilot use third-party knowledge base articles?

No. Copilot only references KBAs stored inside Atera’s native knowledge base. To use third-party articles, you must import or rewrite them into your Atera KBA library.

What happens if domain sync fails in Atera?

If domain sync fails:

• User and device data from Active Directory won’t update
• Ticket assignment may be inaccurate due to stale user records
• You may receive notifications or automation errors
• Previously synced data is retained but may be outdated
   

To resolve:

• Check Admin > Directory Sync
• Validate credentials and network access
• Review logs or contact Atera Support

Best Practice: Enable alerts or regular checks to avoid long sync gaps.

Do I need to retrain Atera Copilot after adding KBAs or scripts?

No manual retraining is required. Once you implement a KBA or script via Actionable Insights, Atera Copilot automatically incorporates it into future suggestions and remediation flows.

Is feedback submitted via live chat used to improve Atera AI Copilot?

Yes. Feedback sent through the Help menu—whether via live chat or idea submission—is logged and fed into Atera’s AI training pipeline. It helps guide product updates and improves the relevance of Copilot’s future suggestions.

How does assigning agents and syncing contacts improve Copilot’s responses?

Assigning agents and syncing contact info ensures:

• Correct technician context when Copilot suggests or triggers actions
• SMS features (e.g., password resets) function properly with valid phone numbers
• User identity and escalation paths are accurate for each client
  This leads to smarter, more context-aware AI behavior.

Can I monitor how well Copilot is trained or optimized in Atera?

Yes. The Optimization Status in AI Center shows:

• Overall completion rate for your site
• Client-level readiness (for MSPs)
  Review which KBAs, scripts, agent assignments, or integrations are missing to fine-tune Copilot’s capabilities.

What’s the best way to start training Copilot for personalized responses?

Begin by:

1. Visiting AI Center > Actionable Insights
2. Reviewing and implementing suggested KBAs and scripts
3. Completing the AI Actions setup (agent assignments, contact sync, domain connections) under Optimize

This gives Copilot the context it needs to tailor replies to your workflows.