FAQs on AI-based support assistance
Site24x7 uses AI-assisted support to help support engineers investigate and resolve support requests more efficiently and consistently. The support engineer provides the runbooks from different modules, diagnostic conditions, and relevant diagnostic information, including log files. The Support Debug agent analyzes the issue using this information and organizes its findings for review by the support engineer.
The agent's role is limited to assisting with the investigation. It does not replace the support engineer's judgment or independently resolve support requests. All findings are reviewed by a support engineer before any conclusions or further actions are taken.
How an investigation works
A typical AI-assisted investigation involves the following steps:
- The support engineer reviews the reported issue and provides relevant details or diagnostic files, when required, to the Support Debug agent.
- The Support Debug agent identifies the applicable user-provided diagnostic runbooks and coordinates the available diagnostic checks.
- The agent presents the progress and results of the investigation, along with supporting evidence where available.
- The support engineer reviews the findings, determines whether further investigation is required, and decides on the appropriate response or follow-up action.
The agent helps select and coordinate diagnostic checks and summarizes their results. While some checks follow predefined diagnostic procedures, agent-generated interpretations are not guaranteed diagnoses. If the available information is insufficient, the agent may request additional information, or the support engineer may continue the investigation manually.
Information and privacy
Information relevant to an investigation, including the issue description, diagnostic results, and uploaded log files, may be processed by the support service and the AI service configured for the deployment.
Uploaded log files are retained temporarily and automatically expire based on the service's retention settings. The default retention period for uploaded log files is six hours. Investigation conversations and related results may also be retained by the support service for operational purposes. For information about the retention and deletion policies applicable to your deployment, contact the support team.
Automated guardrails are used to mask selected sensitive patterns in diagnostic information.
When sharing diagnostic information, avoid sharing passwords, access tokens, or other secrets in uploaded files.
During the investigation, the Support Debug agent verifies customer consent before starting the investigation.
- If consent has not been provided, the agent throws an error message indicating that customer consent has not been provided and remains disabled until consent is enabled in the configuration. In such cases, the support engineer continues the investigation manually.
- If the consent is provided, the agent proceeds with the investigation.
Safeguards and human review
The following table summarizes the responsibilities of the Support Debug agent and the support engineer.
Action | Performed by |
Problem statement initial assessment | Support engineer |
Forwarding the problem for investigation to Support Debug agent | Support engineer |
Gather diagnostic information | Support Debug agent |
Analyze diagnostic information | Support Debug agent |
Organize diagnostic findings | Support Debug agent |
Suggest fixes | Support engineer |
Update support requests | Support engineer |
Close support requests | Support engineer |
AI-generated findings may be incomplete or incorrect. Support engineers review the available evidence and use their judgment before sharing conclusions or taking further action. Based on the reported issue and the relevant information provided by the support engineer, the Support Debug agent performs the applicable diagnostic checks and analyzes the available information to generate its findings.
Frequently Asked Questions
Here are answers to some commonly asked questions about AI-based support assistance.
- What data is shared with AI for support assistance?
Only the configuration, performance, and log data relevant to debugging your support request is shared with OpenAI. All data shared with AI is anonymized and sensitive information is masked. - Is any sensitive data shared with AI?
No. Sensitive information such as email addresses, phone numbers, passwords, API keys, device keys, tokens, and personal identifiers is masked before it is shared with AI. - Where is the data processed?
The shared data is processed by OpenAI in the United States. - Will this data be used to train AI models?
No. The shared data is used only for the specific support or debugging task requested and is not used to train or improve AI models by Site24x7 or OpenAI. - How is the data protected when shared with AI?
Sensitive data is masked before transmission and guardrails are in place to ensure that the data is used only for its intended support purpose. - Can I withdraw my consent?
Yes. You can withdraw your consent anytime by going to Admin > My Account. In the I agree to share my data with AI for support assistance field, select No to disable this option. - What happens if I don't consent to AI-assisted support?
Your support request can still be handled through the default support process without sharing your configuration data with AI.