1
Identify
Understand the customer’s issue and account context.
Meet your AI Front Office
AI Customer Support
Resolve routine customer questions and send exceptions to the right person with the conversation already summarized.
💬
AI Customer SupportRoutine help and accountable escalation
MissionResolve routine questions quickly and send exceptions to the right person with context.
StatusSupports existing customers
Connected toolsKnowledge Base, account context, tickets, routing
Human handoffComplaints, refunds, safety concerns, or policy exceptions
Responsibilities
What the AI Customer Support is responsible for.
2
Resolve
Use verified support knowledge and approved actions.
3
Escalate
Create a clear handoff when human judgment is needed.
Workflow
How work moves through the role.
CustomerReports an issue
SupportChecks context
ResolutionApproved help
EscalationHuman if needed
RecordOutcome saved
Business value
Designed to improve the customer journey—not just answer messages.
Faster help
Routine customers do not wait for simple answers.
Better handoff
Staff receive the issue, history, and actions already attempted.
Consistent policy
Support answers come from the same approved knowledge.
Guardrails
Clear boundaries and accountable human control.
Designed to handle
- Routine account, schedule, service, and policy questions
- Resending confirmations or approved instructions
- Support-ticket creation and status updates
Requires a person
- Refunds, disputes, threats, or complaints requiring judgment
- Security, identity, or privacy incidents
- Safety concerns and policy exceptions
FAQ
Questions businesses usually ask.
Does this role work independently?
No. It works inside the Sochalit AI Front Office and shares approved knowledge, customer context, workflows, permissions, and handoff rules.
Can the role take actions automatically?
Only actions approved during implementation. Sensitive actions can require human approval or be prohibited entirely.
What happens when it does not know?
It should explain that it needs help, capture the question and contact information, and route the inquiry for human review instead of guessing.
Can the role be customized?
Responsibilities, tone, tools, qualification rules, knowledge access, and escalation are configured around the business, while Sochalit’s underlying orchestration and prompt architecture remain proprietary.
See it in context
A conversation should lead somewhere useful.
These examples demonstrate the intended behavior. Final conversations depend on the business’s approved knowledge, rules, integrations, and escalation policies.
Example conversationIllustrative workflow
● Live
C
I cannot find the confirmation email.10:00 AM
S
I can resend it. Before I do, please confirm the email address used for the appointment.10:02 AM
C
It is customer@example.com.10:04 AM
S
Thank you. The confirmation has been resent. If it does not arrive, I will notify the team to help.10:06 AM
Works as one team