Customer Service
AI handles questions with answers in your knowledge base or your business systems: pricing, services offered, service area, hours, availability, appointment status, invoice details. It doesn't handle questions requiring judgement, access to information it doesn't have, or genuinely novel situations. About 70–80% of customer questions fall in the first category.
Specific question types AI handles well: 'How much does [service] cost?' — answered from your pricing FAQ. 'Do you service [city]?' — answered from your service area configuration. 'When is my appointment?' — answered by looking up the customer's booking. 'Can I reschedule?' — answered by checking availability and offering options. 'What's on my invoice?' — answered from the invoice record. 'How long does [service] take?' — answered from your service descriptions.
Question types AI doesn't handle well: 'Should I repair or replace my [equipment]?' — needs professional assessment. 'Why did you charge me this amount?' — may involve a dispute requiring human judgment. 'Can you give me a discount?' — policy and negotiation. 'I have a complaint about [specific technician]' — needs HR or management involvement. These get escalated.
The knowledge base quality is the binding constraint. An AI with a comprehensive, accurate knowledge base can handle a much wider range of questions than one with sparse or outdated information. Building the knowledge base during setup — your pricing, your policies, your services, your FAQs — determines how much question volume the AI resolves without human involvement.
Common questions
The AI acknowledges it doesn't have that information and offers to connect the customer with a human who can help. It doesn't make up an answer. Unanswered questions are logged — a useful source of information for expanding your knowledge base.
Get Started
Tell us what you're working on and we'll show you exactly where automation fits.