Skip to main content

AI Agents

Deploy AI Agents to automate your sales workflows. From responding to website visitors to qualifying leads, AI Agents work 24/7 to drive revenue.

What are AI Agents?​

AI Agents are intelligent assistants that handle sales tasks automatically. They can chat with website visitors, qualify leads, book meetings, and manage outreach - all without human intervention.

Agent Types​

AgentPurpose
Chat AgentEngage website visitors in real-time
Email AgentHandle email replies and follow-ups
LinkedIn AgentManage LinkedIn outreach and responses
Qualification AgentScore and qualify inbound leads
Booking AgentSchedule meetings automatically

Chat Agent​

Setup Chat Agent​

  1. Go to Agents → Chat Agent
  2. Click Create Agent
  3. Configure settings:
    • Agent name and avatar
    • Greeting message
    • Working hours
    • Handoff rules

Chat Widget​

Add the chat widget to your website:

<script>
  window.salestoolsSettings = {
    agentId: 'YOUR_AGENT_ID',
    position: 'bottom-right',
    primaryColor: '#0066FF'
  };
</script>
<script src="https://widget.salestools.io/chat.js" async></script>

Chat Capabilities​

The Chat Agent can:

  • Answer product questions
  • Qualify visitors with questions
  • Book meetings directly in chat
  • Capture lead information
  • Hand off to human agents
  • Support multiple languages

Training Your Chat Agent​

  1. Go to Agents → Chat Agent → Training
  2. Add knowledge:
    • Product FAQs
    • Pricing information
    • Company details
    • Competitor comparisons
  3. Test conversations
  4. Review and improve responses

Email Agent​

How Email Agent Works​

The Email Agent handles your outbound email responses:

  1. Monitors connected email accounts
  2. Classifies incoming replies
  3. Generates appropriate responses
  4. Sends or queues for review

Configuration​

  1. Go to Agents → Email Agent
  2. Enable for specific campaigns
  3. Configure:
    • Response tone
    • Objection handlers
    • Meeting booking integration
    • Human handoff triggers

See AI SDR for detailed setup.

LinkedIn Agent​

LinkedIn Automation​

The LinkedIn Agent handles:

  • Connection requests
  • Welcome messages
  • Follow-up sequences
  • Response handling

Setup​

  1. Go to Agents → LinkedIn Agent
  2. Connect your LinkedIn account
  3. Configure limits and timing
  4. Create message sequences

See LinkedIn Integration for details.

Qualification Agent​

Lead Scoring​

Automatically score leads based on:

FactorPoints
Company size match+20
Industry match+15
Title/role match+15
Engagement level+10-25
Budget indicated+20
Timeline urgency+15

Qualification Questions​

Configure questions for lead qualification:

  1. Go to Agents → Qualification
  2. Add qualifying questions:
    • What's your company size?
    • What problem are you trying to solve?
    • What's your timeline?
    • Who else is involved in the decision?
  3. Set scoring rules for answers
  4. Define qualification thresholds

Routing Rules​

Route qualified leads automatically:

  • Score 80+ → Enterprise sales
  • Score 50-79 → Mid-market sales
  • Score under 50 → Nurture sequence

Booking Agent​

Calendar Integration​

Connect your calendar:

  1. Go to Settings → Integrations
  2. Connect Calendly, Cal.com, or Google Calendar
  3. Configure availability
  4. Set meeting types

Booking Flow​

The Booking Agent:

  1. Identifies meeting intent
  2. Offers available times
  3. Confirms timezone
  4. Creates calendar event
  5. Sends confirmation email
  6. Adds reminder sequence

Meeting Types​

Configure different meeting types:

TypeDurationDescription
Discovery Call30 minInitial qualification
Demo45 minProduct demonstration
Technical Call60 minDeep dive with engineering
Follow-up15 minQuick check-in

Agent Settings​

General Settings​

SettingDescription
NameHow the agent introduces itself
AvatarProfile image for chat
ToneProfessional, friendly, casual
LanguagePrimary language + translations
Working HoursWhen agent is active

Handoff Rules​

Configure when to involve humans:

  1. Keyword triggers - Specific words route to human
  2. Sentiment - Negative sentiment = handoff
  3. Request - "Speak to human" phrases
  4. Complexity - Questions AI can't answer
  5. VIP accounts - Key accounts get humans

Notifications​

Set up notifications:

  • New conversations
  • Handoff requests
  • High-value leads
  • Booking confirmations

Analytics​

Agent Performance​

Track agent effectiveness:

MetricDescription
ConversationsTotal interactions
Resolution Rate% handled without handoff
Response TimeAverage reply speed
SatisfactionUser ratings
Meetings BookedSuccessful bookings
Leads QualifiedQualification rate

Conversation Review​

Review agent conversations:

  1. Go to Agents → Conversations
  2. Filter by agent, date, or outcome
  3. Review transcripts
  4. Provide feedback for training

Best Practices​

Agent Training​

  1. Start with FAQs - Add common questions first
  2. Review conversations - Learn from real interactions
  3. Iterate regularly - Update training weekly
  4. Test thoroughly - Try edge cases before launch

Maximizing Effectiveness​

  1. Clear handoff rules - Know when humans should step in
  2. Fast response - Speed matters in chat
  3. Personalization - Use visitor data when available
  4. Follow-up - Don't let conversations die

Common Mistakes​

  1. Over-automating - Some situations need humans
  2. Generic responses - Personalization matters
  3. Ignoring feedback - Learn from poor interactions
  4. Missing handoffs - Frustrates visitors

Troubleshooting​

Agent Not Responding​

  • Check agent is enabled
  • Verify working hours
  • Review trigger conditions
  • Check integration connections

Poor Response Quality​

  • Add more training data
  • Review and correct responses
  • Adjust tone settings
  • Add specific use cases

Missed Handoffs​

  • Review handoff rules
  • Add keyword triggers
  • Lower confidence threshold
  • Test with edge cases

Support​

Need Help?

Last Updated: January 2025 Version: 1.0