Why isn’t AI responding?

Possible causes include:

• Temporary connectivity issues

• AI permissions not enabled for your account

• AI token availability limits

• Large requests still processing

Try simplifying your request or contact your RMI Client Technology Team if the issue continues.

Can Administrators customize AI prompts?

Depending on your organization's configuration, Administrators may create reusable AI prompts to standardize analyses, summaries, and reports across the organization using a Google Drive or equivalent.

How is AI data protected?

SafetyOne™ AI operates within your organization's security boundaries and respects user permissions by leveraging private agents.

AI responses are generated using only the information users are authorized to access.

Your AI interactions are not shared with other organizations using SafetyOne.

How accurate are AI responses?

SafetyOne™ AI is designed to provide helpful summaries and recommendations based on the information available.

Training recommendations from SafetyOne AI should always be checked to ensure they are following local, state, and federal guidelines.

Does AI replace safety professionals?

No.

SafetyOne™ AI is designed to assist—not replace—human expertise.

AI can quickly identify trends, summarize information, and recommend actions, but important operational and safety decisions should always be reviewed by qualified personnel.

Who can use AI?

AI permissions are only granted to Administrators.

How can I get more AI tokens?

AI tokens can be purchased by contacting your Rancho Mesa Client Technology Team. Tokens will be posted to your account within one to two business days following the completion of purchase.

What are AI Tokens?

AI Tokens are units used to process AI requests.

Different requests consume different numbers of tokens depending on factors such as:

• Amount of information analyzed

• Complexity of the request

• Size of attached documents

• Length of AI responses

Your organization manages AI token availability and usage.

Token count can be viewed by administrators at the top of the menu panel on the SafetyOne web browser.

Are my AI conversations private?

Yes.

AI responses are generated using your organization's authorized data and respect user permissions.

Administrators only receive information they are authorized to access.

Your AI interactions are not shared with other organizations using SafetyOne.

Does AI learn from my company's data?

SafetyOne™ AI analyzes the information available within your organization's SafetyOne dashboard during each request to generate relevant responses.

It does not permanently "learn" from individual conversations in a way that changes future behavior without your organization's control.

AI responses are based on the data and permissions available at the time of the request.

Can I ask AI questions in plain English?

Yes.

SafetyOne™ AI supports natural language questions such as:

• "What hazards are increasing this month?"

• "Which project has the most Open Issues?"

• "Summarize this inspection."

• "Show me recurring forklift problems."

• "What training is overdue?"

• "Generate a Toolbox Talk from today's observations."

No special commands or technical knowledge are required.

Can supervisors edit AI-generated Toolbox Talks?

Yes.

AI-generated content is intended to assist—not replace—your safety professionals.

Supervisors can review, edit, expand, or customize any Toolbox Talk before adding it to the platform and distributing it to employees.

Can AI use observations to generate Toolbox Talks?

Yes.

SafetyOne AI can analyze hundreds or thousands of safety observations and identify recurring themes.

It can then generate Toolbox Talks addressing the hazards employees are encountering rather than relying solely on generic industry content.

Can AI generate Toolbox Talks from incidents?

Yes.

AI can review incident reports and generate Toolbox Talks focused on:

• Root causes

• Lessons learned

• Prevention strategies

• Corrective actions

• Similar hazards

This helps organizations immediately reinforce learning after an incident occurs.

Can AI generate Toolbox Talks?

Yes.

SafetyOne™ AI can create customized Toolbox Talks using information from your organization's own data.

Instead of generic safety topics, AI can generate discussions based on actual conditions affecting your workforce.

Can AI recommend additional training?

Yes.

SafetyOne™ AI can identify trends within observations, inspections or incidents and recommend training topics that may help reduce future risk.

Recommendations assist managers in making informed training decisions but should always be reviewed before assignment.

What kinds of information can AI analyze?

SafetyOne AI analyzes structured information stored within your SafetyOne environment, including:

• Safety inspections

• Incident reports

• Near misses

• Equipment inspections

• Environmental inspections 

• Corrective actions

• Historical trends

What is SafetyOne™ AI Data Analytics?

Rancho Mesa now offers GPT-5.1-powered SafetyOne™ AI analysis capabilities across the following modules: Observations and Mobile Forms. AI analysis for Toolbox Talks and Trainings will be available in future releases.

The AI Analysis is embedded directly inside your SafetyOne Administrator website to help you:

• Summarize large volumes of reports

• Identify recurring hazards and trends

• Draft toolbox talks from real jobsite data

• Generate executive summaries

• Highlight root causes

• Turn structured tables into readable narratives.