You can talk to AI using normal, everyday language. You do not need to learn commands, field names, or a special syntax. Simply describe what you want to know or what you want to do, much as you would explain it to a colleague.
For example:
- What is our forecast for this month?
- Show me the deals that have been inactive for more than 30 days.
- Create a contact for Marco Rossi.
- Schedule a meeting with Carlo tomorrow at 3 PM.
- Create a quote for Carlo for 10 licenses of Product X.
- Send this quote for signature.
Be as specific as you need
The more useful context you provide, the easier it is for onAI to understand exactly what you want.
For example:
instead of: Find the deal.
you can say: Find the open deal for ABC Consulting.
and instead of: Create a meeting.
you can say: Schedule a 30-minute meeting with Carlo tomorrow at 3 PM.
You do not need to provide every detail immediately. If information required to complete the request is missing or ambiguous, onAI may ask you a follow-up question before continuing.
You can continue with follow-up questions
You do not need to repeat the entire request every time.
After asking: What is our forecast for this quarter?
you can continue with:
- What is still open?
- Which deals are contributing the most?
- Show me only the deals expected to close this month.
onAI uses the conversation context to understand follow-up requests when the reference is clear.
The same applies to operational workflows. For example: Create a quote for Carlo.
onAI may ask for information that is still required. You can simply answer the question and continue the workflow.
You can ask for more than one related action
When several actions belong to the same objective, you can describe them together.
For example: Create a quote for Carlo and send it for signature.
or: Create the quote and prepare an email for Carlo.
onAI can coordinate supported actions as part of the same workflow when they are related to the same objective.
You can refer to the current context
When the context is clear, you can use natural references such as: this deal, this contact, that quote, send it for signature, move it to tomorrow.
If there is more than one possible record or the reference is ambiguous, onAI may ask you to clarify which one you mean.
Ask naturally
You do not need to know the exact database field, internal status name, or technical structure used by Onpipeline.
For example, you can ask: Which opportunities haven’t had any activity recently?
rather than trying to specify the exact fields or filters required to calculate it.
onAI translates the request into the supported CRM operations needed to answer the question.
Review actions before confirming them
When a request will change CRM data or execute a supported operation, onAI can prepare the action first and show you what is going to happen. Review the proposal and choose Confirm when you want the operation to proceed, or Cancel if you do not want to apply it.
If something in the proposal is not what you intended, you can cancel it and explain the change conversationally.
For example: Use 20 licenses instead of 10.
or: Schedule it for Friday instead.
If onAI needs clarification
Sometimes a name, date, record, or instruction can refer to more than one thing.
In those cases, onAI may ask a short clarification question instead of making an assumption.
For example: Which Carlo do you mean?
or: Which team should be used for this quote?
Answer normally and onAI will continue from there.
If a capability is not available
If you ask for something that is not currently supported by AI, it will tell you that the capability is not yet available and indicate the areas it can currently work with.
Talk to onAI as you would talk to a colleague. Tell it the outcome you want; you do not need to tell it how to navigate the CRM.