CallVIQ – AI Voice Calling Automation & CRM SaaS (Node.js script)

CallVIQ is a self-hosted, multi-tenant AI voice calling platform built for agencies, entrepreneurs, and SaaS resellers. It replaces expensive human call centers with AI voice agents that handle outbound marketing calls, lead qualification, appointment scheduling, and customer support around the clock. Buyers install it on their own server, white-label it completely, and resell AI calling services to their own clients under their own brand.

Built on a Bring Your Own Key (BYOK) model, CallVIQ routes all API usage costs directly to each tenant’s personal billing accounts across Twilio, OpenAI, Gemini, ElevenLabs, and other connected providers. The platform host never carries usage costs on behalf of tenants. Each third-party service is billed independently by its provider at its own rate — please review the Third-Party API Costs section below before purchasing.

Demo Credentials

A live demo is available for preview before purchase. Use the credentials below to explore the platform.

Super Admin Panel: https://callviq.ncodeslab.com/superadmin/login
Email: admin@callviq.com
Password: Admin@123456

Tenant Panel: https://callviq.ncodeslab.com/
Email: demo@callviq.com
Password: Password@123!

Documentation: Here

Key Features at a Glance

• Multi-tenant workspace architecture with complete white-label custom domain and branding support
• 5 Supported Telephony Providers: Native integration with Twilio, Telnyx, Plivo, Vonage Voice API, and custom SIP Trunking
• 9 Supported AI LLM Providers: Google Gemini (default: Gemini 2.0 Flash), OpenAI / ChatGPT (GPT-4o, o3-mini), Anthropic Claude (3.5 Sonnet, 3.5 Haiku), Groq LPU, Mistral AI, xAI (Grok-2), DeepSeek API, OpenRouter, and Cerebras API
• 9 Supported Voice & Speech Synthesis Engines: ElevenLabs (Turbo v2.5 & Indian accent presets), Cartesia AI, OpenAI Voice, Vapi Voice Agent Gateway, Microsoft Azure Neural TTS, Deepgram Aura TTS, Google Cloud TTS, AWS Amazon Polly, and Microsoft Edge Free TTS
• Drag-and-drop visual conversation flow builder with 7 node types, prompt context substitution, fallback prompts, and live human call transfers
• RAG Knowledge Base engine with PDF parsing, TXT/CSV indexing, and automated web URL crawler with MySQL Full-Text search context injection
• Enterprise CRM Integration Hub with bi-directional contact sync for HubSpot CRM, Salesforce Cloud, Zoho CRM, and Google Sheets, plus Kanban lead pipelines and 100k row CSV import wizard
• Google Workspace integration with staff resource binding, Google Calendar real-time slot availability checking, automated Google Meet link generation, and local database offline fallback
• Built-in Graphics & Visual Studio canvas for generating marketing banners, templates, and social collateral inside the application
• Stripe subscription billing engine with credit ledger, dynamic plan tier overrides, auto-recharge thresholds, and real-time webhook lifecycle
• Live Call Monitoring dashboard with WebSocket state push updates, silent listening, force call termination, diarized transcripts, AI summaries, and lead scoring (0 to 100)
• Multi-tenant Support Ticketing Portal with ticket categories, priority levels, and file attachment support
• Full Super Admin Panel with 4-tier RBAC (Owner, Administrator, Support Agent, Viewer), tenant impersonation, provider master toggles, and system health telemetry
• In-browser call simulator testing playground with hands-free Voice Activity Detection (VAD) loop and live transcript inspection
• Global and tenant-level Do-Not-Call (DNC) registry pre-screening for total regulatory compliance
• AES-256-GCM key-at-rest encryption for all tenant BYOK API keys, SIP secrets, and OAuth tokens
• One-command Docker Compose deployment stack

Super Admin Panel

The Super Admin console gives platform owners total administrative control. Create and manage tenant accounts manually or via automated email invitations. Suspend, reactivate, or soft-delete tenants with a single click. Impersonate any tenant workspace for immediate support and troubleshooting. Build subscription plans dynamically, set custom Stripe price IDs, set per-plan feature flags (inbound calling, CRM integrations, call recording), overage rates, and free trial allocations.

Platform settings cover full white-label branding (logo, platform name, primary/accent colors, custom domain headers), mailer configuration (Nodemailer SMTP or Resend API), and global master toggles for each of the 5 Telephony providers, 9 AI LLM providers, 9 Voice engines, and 4 CRM integrations. The built-in legal page editor manages Terms of Service, Privacy Policy, DNC compliance guidelines, and system notification banners served dynamically to the application.

The analytics dashboard tracks MRR, ARR, active subscriber growth, total call minutes, LLM token throughput, and credit ledger operations. A real-time system health monitor displays live CPU, RAM, disk usage, MySQL connection pool status, Redis queue depths, and API connection latency for all third-party provider integrations.

Super Admin access uses a 4-tier Role-Based Access Control (RBAC) model: Owner, Administrator, Support Agent, and Viewer, enforced by safety locks that prevent owner orphan states and unauthorized privilege escalation.

Tenant Dashboard

Tenants are guided through an interactive multi-step onboarding wizard covering business profile creation, plan tier selection, telephony provider credential hookup (Twilio, Telnyx, Plivo, Vonage, or custom SIP), and BYOK AI and voice key configuration. All API keys and secrets are encrypted at rest using AES-256-GCM.

The AI Agent Builder lets tenants construct voice agents using a visual drag-and-drop node graph. Supported node types include Greeting, Question, Condition (Yes/No branching), Transfer to Human, Book Appointment, Knowledge Retrieval (RAG lookup), and End Call. System prompts support dynamic bracket variable substitution such as {first_name}, {company}, or {order_id}. Agents can be cloned, configured with custom LLM temperature and token limits, assigned specific voice models, and linked to a RAG Knowledge Base built from uploaded documents or crawled URLs.

The Campaign Manager supports five campaign formats: Outbound Blast, Drip Sequence, Event-Triggered, Inbound Dynamic Routing, and Appointment Reminders. Dialer rules include phone number rotation, calling-hour window enforcement, retry limits, call frequency caps, and automated credit depletion safety pauses.

The Live Call Monitor streams all active calls in real time via WebSocket push updates. Admins can listen in silently or terminate calls directly from the dashboard interface. Post-call processing populates full diarized transcripts, AI-generated summaries, sentiment scores (positive, neutral, negative), outcome tags, and lead scores (0 to 100).

The CRM Module manages contacts with up to 20 custom fields, tag-based segmentation, and a Kanban lead pipeline with drag-and-drop status updates. Contact lists support CSV imports of up to 100,000 rows with column mapping and duplicate handling rules (skip vs. update).

The Testing Playground provides an in-browser call simulator with a visual mobile interface, live status indicators, speech-to-text response previews, and a hands-free Voice Activity Detection (VAD) loop for testing agents before going live.

The Support Ticketing System allows tenants to submit support tickets (Billing, Telephony, Technical, General) with priority levels, base64 file attachments, and interactive discussion threads with platform admins.

Real-Time Calling Engine & Telephony Layer

CallVIQ V2.0 operates a low-latency telephony pipeline built on bidirectional WebSockets. When a call connects, BullMQ pops contacts from the dialing queue, initiates outbound calls via cloud telephony REST APIs, and connects a real-time WebSocket audio stream (wss://) streaming raw base64-encoded μ-law audio chunks directly to the backend.

Supported Telephony Providers:

• Twilio: Twilio REST API + Twilio Media Streams audio WebSocket
• Telnyx: Telnyx Voice API + TeXML / Call Control WebSocket streams
• Plivo: Plivo Voice API + Stream XML audio channel
• Vonage (Nexmo): Vonage Voice API + NCCO WebSocket audio streams
• SIP Trunking: Generic SIP trunks (Asterisk, FreePBX, Kamailio, OpenSIPS) for direct PBX integration

Audio is bridged in real time to the selected AI provider — Gemini Live API or OpenAI Realtime for direct streaming pipelines, or a chained STT (Whisper / Deepgram), LLM inference, and TTS pipeline for non-streaming setups. The engine includes sub-second barge-in interruption handling, silence detection with fallback prompts, and post-call job dispatching for transcript generation, CRM updates, and outbound webhooks.

For inbound calls, the system maps the caller’s phone number to an existing CRM contact record and preloads prompt variables before starting the AI session.

AI & LLM Orchestration Layer

CallVIQ V2.0 dynamically routes conversation prompts through a multi-provider LLM orchestration engine. Tenants can supply custom API keys or rely on platform defaults across 9 supported AI LLM providers:

• Google Gemini (Default API): Gemini 2.0 Flash (ultra-fast latency), Gemini 1.5 Pro, and Gemini 1.5 Flash
• OpenAI / ChatGPT: GPT-4o, GPT-4o Mini, o3-mini (reasoning model), and GPT-4 Turbo
• Anthropic Claude: Claude 3.5 Sonnet, Claude 3.5 Haiku (conversational model), and Claude 3 Opus
• Groq LPU Engine: Llama 3.3 70B Versatile and Llama 3.1 8B Instant (ultra-fast LPU inference)
• Mistral AI: Mistral Large Latest and Open-Mixtral 8×22B
• xAI (Grok): Grok-2 Latest OpenAI-compatible API
• DeepSeek API: DeepSeek Chat and DeepSeek Reasoner high-performance inference
• OpenRouter API: Unified cloud prompt routing engine
• Cerebras API: Cerebras ultra-fast Llama 3.3 70B inference

Voice & Speech Synthesis Engine

CallVIQ V2.0 streams text-to-speech audio synthesis across 9 supported voice provider options:

• ElevenLabs: High-fidelity audio synthesis using the eleven_turbo_v2_5 engine with stability and similarity controls. Includes American voice presets (Rachel, Domi, Bella, Antoni, Adam, Nicole), Indian accent presets (Ananya, Rohan, Kavya, Aarav, Aditi, Dev), and Multilingual presets (Clyde, Dave, Freya, Giovanni).
• Cartesia AI: Ultra-low latency streaming TTS using the sonic-english model (voices: Blake, Jessica, Chloe, Austin).
• OpenAI Voice: Low-latency tts-1 engine (voices: alloy, echo, fable, onyx, nova, shimmer).
• Vapi Voice Agent Gateway: Proxied end-to-end voice-agent wrapper for direct telephony proxying.
• Microsoft Azure Cognitive Services: Neural speech engine (default: en-US-JennyNeural).
• Deepgram Aura TTS: Low-latency neural TTS (default: aura-asteria-en) + Deepgram STT.
• Google Cloud Text-to-Speech: Neural2 speech synthesis (default: en-US-Neural2-A).
• AWS Amazon Polly: Cloud speech generation (default: Joanna).
• Microsoft Edge Free TTS: Built-in zero-cost fallback speech engine (default: en-US-AvaNeural).

Google Workspace & Enterprise CRM Integration Hub

Google Workspace: Tenants connect their Google accounts to enable calendar-aware appointment booking. The AI agent verifies slot availability against both local reservations and staff Google Calendars before confirming meetings. Confirmed bookings create Google Calendar events complete with location details and Google Meet video URLs. If Google Calendar is not connected, conflict checking automatically runs against the local database so booking never breaks. Google Sheets integration provides two-way contact syncing for import and export workflows.

Enterprise CRM Integration Hub: Native 2-way connectors for HubSpot CRM, Salesforce Cloud, and Zoho CRM. Incoming caller phone numbers are mapped to CRM contacts in real time, and call outcome tags, diarized transcripts, and AI summaries are indexed directly onto CRM contact timelines.

Billing and Credit System

CallVIQ integrates Stripe for subscription billing. Tenants checkout, manage payment methods, download invoices, purchase credit top-ups, and configure auto-recharge thresholds through Stripe-hosted portals. The platform processes the full Stripe webhook lifecycle including subscription creation, plan upgrades with proration, cancellations, and payment failures. Credits are tracked via Redis atomic operations. When a tenant’s credit balance reaches zero, active calls complete gracefully but new calls are blocked until credits are refilled.

Security and Compliance

All tenant API credentials — including Twilio SIDs, Telnyx keys, Plivo auth tokens, Vonage keys, SIP parameters, OpenAI/Gemini/Anthropic keys, ElevenLabs keys, and Google OAuth tokens — are encrypted at rest using AES-256-GCM. GDPR compliance tools include a complete data-dump export generating a ZIP package of all tenant records in JSON and CSV formats, as well as a hard-delete option that permanently purges contact data. The calling engine automatically enforces calling hour restrictions, call frequency caps, and consent notice playback before recording starts. Global and tenant-level DNC blocklists screen numbers before dialing.

Tech Stack

• Frontend: Next.js 16.2.9 (App Router, Turbopack), React 19, Tailwind CSS v4, Lucide Icons, Sonner Toasts, Radix UI
• Backend API: Node.js, Express.js, TypeScript
• Real-Time Engine: Node.js WebSockets (ws)
• Job Queue: BullMQ + Redis 7
• Database: MySQL 8.0 / MariaDB 10.4 (42 normalized tables with connection pooling)
• Cache: Redis 7 for session tokens, rate limiting, and active call state
• File Storage: AWS S3, Backblaze B2, or MinIO
• Mailer: Nodemailer SMTP or Resend API
• Payments: Stripe SDK
• Telephony: Twilio REST/Streams, Telnyx Voice API, Plivo Voice API, Vonage Voice API, Custom SIP Trunks

What Is Included

• Complete Next.js 16 frontend source code (App Router + Turbopack)
• Complete Node.js / Express TypeScript backend API source code
• Node.js WebSocket real-time voice call engine
• Full MySQL 8.0 database schema (42 normalized tables + seed dataset)
• WhatsFlow WhatsApp messaging sub-platform source code
• Integrated Visual Banner & Graphics Studio source code
• Docker Compose stack for one-command deployment
• Environment validator setup script
• Annotated .env.example configuration templates
• Admin installation guide (PDF / Markdown)
• Tenant onboarding guide (PDF / Markdown)
• 6 months of item support and updates

Third-Party Services

This platform integrates third-party services that are billed separately by their respective providers and are not included in the purchase price. Each service is connected using your own account credentials (BYOK). Costs vary by provider, usage volume, and selected plan. The following services are supported:

• Telephony Providers (5): Twilio, Telnyx, Plivo, Vonage Voice API, and Generic SIP Trunking
• AI LLM Providers (9): Google Gemini, OpenAI (ChatGPT), Anthropic Claude, Groq LPU, Mistral AI, xAI (Grok), DeepSeek, OpenRouter, and Cerebras
• Voice & Speech Providers (9): ElevenLabs, Cartesia AI, OpenAI Voice, Vapi Voice Gateway, Microsoft Azure Neural TTS, Deepgram Aura TTS & STT, Google Cloud TTS, AWS Amazon Polly, and Microsoft Edge Free TTS
• CRM Integrations (4): HubSpot CRM, Salesforce Cloud, Zoho CRM, and Google Sheets API
• Workspace Tools: Google Calendar API and Google Meet
• Payments & Billing: Stripe SDK
• Storage: AWS S3, Backblaze B2, or MinIO

Third-Party Pricing & Billing

CallVIQ is sold as a self-hosted software license. The purchase price includes only the software source code. Third-party services are NOT included and must be purchased separately from their respective providers under your own BYOK accounts.

Requirements

• Node.js 18.x or 20.x or higher
• MySQL 8.0 or MariaDB 10.4 or higher
• Redis 7 or higher
• A supported telephony account (Twilio, Telnyx, Plivo, Vonage, or SIP provider)
• A Stripe account for billing configuration
• VPS or dedicated server with a minimum of 2 CPU cores and 4 GB RAM recommended
• Docker and Docker Compose (optional, for containerized deployment)

Important Notes

• The purchase includes only the CallVIQ software source code.
• No AI credits are included.
• No telephony credits or phone numbers are included.
• No OpenAI, Gemini, Anthropic, ElevenLabs, Cartesia, or Groq credits are included.
• You must create your own accounts with supported third-party providers.
• Provider pricing may change at any time without notice.
• All third-party fees are paid directly to the respective providers.
• The software vendor is not responsible for third-party pricing or billing.
• An internet connection is required for AI, voice synthesis, and telephony features.

Before Purchasing

Please review the pricing of the third-party telephony, LLM, and TTS services you intend to use before purchasing this item. The total operating cost depends on your selected AI provider, voice synthesis engine, telephony carrier, file storage provider, and monthly calling volume.