Aqib Ali Ghumro
> BACK_TO_SYSTEMS
AI[SYS_ID: AI-CALL-CENTER]

VoicePls / Client CallCenterAI

Voice automation work covering a public VoicePls transcription pipeline and a client CallCenterAI booking agent connected to custom backend workflows.

Open VoicePls

01.System Interface & Screenshots

AI / Backend Engineer
VoicePls / Client CallCenterAI Interface Screenshot
[SCREENSHOT_01_ACTIVE]

02. Project architecture visualization

Voice-agent workflow visualization based on the implemented prototype

A call workflow for booking, rescheduling, cancellation, transfer, and summaries through Vapi tool calls. [This visualization is based on development and test-call scenarios, not live customer traffic.]

Speed:
Caller
Vapi Voice Agent
Speech / Conversation Context
LLM Decision
Tool Call Selection
Custom Booking REST API
Appointment Data Store
Booking Result
Voice Confirmation
Optional SMS / Human Transfer / Summary
Active Pipeline Stage
Launch simulation to view stage details.
Live Stage Data Payload
Data telemetry will load here.
Console Logs
|
>Ready to trace workflow pipelines...

03. The Problem

Businesses need a reliable voice workflow to handle repetitive appointment requests without relying on a human to answer every call.

04. The Solution

Connected Vapi to a custom booking backend, designed the prompt and tool-call flow, and validated appointment actions through backend logic.

05. Key Features

  • >Inbound calls
  • >Outbound calls
  • >Availability checking
  • >Appointment booking
  • >Rescheduling
  • >Cancellation
  • >Human transfer
  • >SMS confirmation
  • >Call recording
  • >Transcripts and call summaries

AI & Automation Workflows

-System prompt design for booking workflows
-Tool and function calls for booking actions
-Fallback logic and error handling for call interruptions

08. Engineering Challenges

Keeping the call flow resilient when callers provide incomplete details

Added fallback prompts and validation before backend actions.

Mapping speech turns to booking tools safely

Separated LLM intent detection from backend execution steps.

[SYSTEM_SPECIFICATIONS]
ROLE:
AI / Backend Engineer
CATEGORY:
AI
TECHNOLOGY STACK:
VapiLLMsJavaScriptREST APIsTool Calling
BUSINESS IMPACT:
Validated through development and test-call scenarios rather than live customer traffic.