During practice sessions, Oliver AI deploys an intelligent call agent that plays the role of a realistic customer. This agent adapts to your responses, raises authentic objections, and provides a training experience that closely mirrors real sales conversations.
When you start a practice session, the AI agent is configured with:
- Customer persona -- A detailed character profile including name, company, role, personality traits, and buying preferences.
- Product knowledge -- Information about your product so the AI can ask relevant questions and raise realistic objections.
- Scenario context -- The type of interaction (cold call, discovery, demo, etc.) which shapes how the AI behaves.
- Company context -- Knowledge from your organization's onboarding data, so the AI understands your industry and competitive landscape.
The AI call agent can:
- Introduce itself -- Start the conversation in character with a name, company, and reason for the call (or respond to your outreach).
- Ask questions -- Inquire about your product, pricing, implementation, and support, just like a real prospect.
- Raise objections -- Push back on your proposals with common concerns such as budget, timing, competition, and need.
- Express emotions -- Show enthusiasm, skepticism, frustration, or confusion depending on how the conversation goes.
- Make decisions -- Agree to next steps, ask for more time, or decline based on your sales approach.
- Remember context -- Maintain conversation history so it references things you discussed earlier in the call.
Each AI customer persona is generated with distinct attributes:
- Geography-aware naming -- Names are culturally appropriate for 15+ regions worldwide.
- Role-specific behaviour -- A C-level executive responds differently than a mid-level manager.
- Industry knowledge -- The persona understands their industry's common challenges and terminology.
- Buying stage -- Awareness of where they are in the buying journey affects their receptiveness.
During the practice session:
- You send a message -- Your text is processed along with the full conversation history and persona context.
- AI responds -- The AI generates a response in character as the customer.
- Response delivered -- The AI customer's reply appears in the chat interface.
- Usage recorded -- Each exchange records AI usage against your organization's current trial or subscription period.
The conversation continues naturally until you or the AI reaches a conclusion.
If your organization has configured a sales coaching framework, the AI agent can also provide real-time coaching signals:
- The framework defines the ideal flow of a sales conversation.
- The system tracks which framework stages you have covered during the call.
- Post-call, the framework alignment is included in the evaluation.
- Be conversational -- Write as you would speak. The AI responds better to natural language than scripted pitches.
- Provide context -- If the AI asks a question, give a detailed answer. Short or vague responses limit the AI's ability to generate realistic follow-ups.
- Challenge yourself -- Do not just agree with everything the AI says. Push for the sale and see how the AI responds.
- Try different approaches -- The AI responds differently based on your strategy. Experiment with consultative selling, challenger selling, and solution selling to see what works best.
Note: The AI agent is designed for training purposes. It simulates realistic customer interactions but does not replicate any specific real person or company.