Creating Your First Agent
Creating an AI voice agent in Recepta.ai is a straightforward process. This guide walks you through each step, from initial setup to going live.Before creating agents, make sure you’ve completed the onboarding process and have your business information configured.
Step 1: Access the Agent Creation Interface
- Navigate to Agents: In your dashboard, click on the “AI Agents” section
- Create New Agent: Click the “Create New Agent” button
- Agent Modal: The agent creation modal will open with configuration options

Step 2: Basic Agent Configuration
Agent Name and Description
Agent Name
Best Practices:
- Use descriptive names (e.g., “Sales Agent”, “Support Bot”)
- Keep it under 50 characters
- Avoid special characters
- Make it easily identifiable
Description
Best Practices:
- Clearly explain the agent’s purpose
- Include primary functions
- Mention target audience
- Keep it concise but informative
Voice Selection
Choose from our library of professional AI voices:Professional Voices
Professional Voices
Best for: Business, corporate, professional services
- Clear articulation
- Neutral accent
- Authoritative tone
- Available in male/female variants
Friendly Voices
Friendly Voices
Best for: Customer service, hospitality, retail
- Warm and approachable
- Conversational tone
- Slight regional accents available
- High customer satisfaction scores
Technical Voices
Technical Voices
Best for: Tech support, software companies
- Clear technical pronunciation
- Patient and helpful tone
- Good for complex explanations
- Popular with B2B companies
Step 3: Define Agent Behavior
General Prompt Configuration
The general prompt is the foundation of your agent’s personality and behavior. It should include:1
Role Definition
Define what role the agent plays in your organization
2
Personality Guidelines
Establish the agent’s communication style
3
Specific Instructions
Include specific behaviors and responses
4
Escalation Rules
Define when and how to escalate calls
Advanced Prompt Examples
- Customer Support
- Sales Agent
- Appointment Scheduler
Step 4: Voice and Language Settings
Language Configuration
Primary Languages
- English (US, UK, AU)
- Spanish (Latin America, Spain)
- French (France, Canada)
- German
- Italian
- Portuguese
Advanced Options
- Accent Selection: Choose regional accents
- Speaking Speed: Adjust words per minute
- Voice Pitch: Modify tone characteristics
- Pause Duration: Control natural pauses
Voice Parameters
Step 5: Integration Setup
Knowledge Base Connection
Connect your agent to relevant knowledge sources:1
Select Knowledge Base
Choose which knowledge bases your agent should access
2
Set Permissions
Define what information the agent can access and share
3
Configure Search
Set up how the agent searches and retrieves information
External Integrations
Calendar Integration
Calendar Integration
Supported Platforms:
- Google Calendar
- Microsoft Outlook
- Calendly
- Acuity Scheduling
CRM Integration
CRM Integration
Supported Platforms:
- Salesforce
- HubSpot
- Pipedrive
- Custom CRM via API
Communication Tools
Communication Tools
Supported Platforms:
- Slack notifications
- Microsoft Teams alerts
- Email notifications
- SMS alerts
Step 6: Testing and Validation
Before activating your agent, thorough testing is crucial:Test Call Process
1
Initial Test
Use the “Test Call” feature in the dashboard to make a test call to your agent
2
Scenario Testing
Test various customer scenarios and edge cases
3
Voice Quality Check
Verify voice clarity, speed, and natural flow
4
Integration Testing
Confirm all integrations (calendar, CRM) are working properly
Test Scenarios Checklist
- Basic Interactions
- Information Requests
- Complex Scenarios
- Agent answers promptly
- Greeting is appropriate
- Agent asks for caller’s name
- Voice quality is clear
- Response time is reasonable
Step 7: Going Live
Once testing is complete, activate your agent:- Final Review: Double-check all configurations
- Set Status to Active: Change agent status from “Training” to “Active”
- Monitor Initial Calls: Watch the first few calls closely
- Collect Feedback: Gather feedback from initial callers
- Iterate and Improve: Make adjustments based on real-world performance
Common Creation Issues
Agent Creation Fails
Agent Creation Fails
Possible Causes:
- Missing required fields
- Invalid prompt formatting
- Voice selection not available
- Integration connection issues
- Check all required fields are completed
- Verify prompt doesn’t contain invalid characters
- Try a different voice option
- Test integrations individually
Voice Quality Issues
Voice Quality Issues
Possible Causes:
- Incorrect voice parameters
- Network connectivity issues
- Server overload
- Adjust speed and pitch settings
- Test from different locations
- Try during off-peak hours
- Contact support if persistent
Integration Problems
Integration Problems
Possible Causes:
- Invalid API credentials
- Insufficient permissions
- Service downtime
- Verify API keys are correct
- Check service permissions
- Test integrations separately
- Review integration logs

