📞

Intelligent Call Transfer

Our AI agent transfers calls with full context in under 800 ms. Forget about customers repeating their problem every time they get transferred: NetVoice transfers the entire history, call summary, and the reason for the transfer directly to the receiving agent, whether human or AI.

<800ms
Transfer Latency
99.9%
Guaranteed Uptime
3 Types
Cold · Warm · Agentic
0%
Context Loss
👤
Cliente: María García
Llamada activa · 02:34 · Problema técnico nivel 2
EN VIVO
🤖 Análisis IA en tiempo real
Detectado: Problema de facturación complejo
Nivel de urgencia: Alto · Sentimiento: Frustrado
Recomendación: Transferencia tibia → Agente especializado
⚡ <800ms
📋 Contexto enviado al agente
Motivo: Cobro duplicado en factura de enero
Historial: 3er contacto · Caso #45821
Resumen: IA verificó cargo, escala a autorización
🧑‍💼
Agente: Carlos Mendoza
Especialista Facturación · Disponible · Contexto recibido ✓
LISTO

Tres modos de transferencia para cada situación

NetVoice ofrece tres tipos de transferencia, cada uno optimizado para distintos escenarios de negocio. El sistema detecta automáticamente cuál es el más apropiado según el contexto de la llamada, la urgencia y la disponibilidad de los agentes.

🔵

Transferencia Fría

La llamada se redirige inmediatamente sin intervención del agente receptor. Ideal para enrutamiento automático en horarios pico cuando la velocidad es prioritaria sobre el contexto.

🟡

Transferencia Tibia

El agente receptor recibe el resumen completo de la conversación antes de conectarse con el cliente. Elimina el 100% de las repeticiones del problema y reduce el tiempo de resolución hasta un 40%.

🟢

Agentic Transfer con Contexto

La IA prepara un briefing completo: historial del cliente, análisis de sentimiento, acciones previas, intenciones detectadas y pasos sugeridos para la resolución. El agente entra sabiendo exactamente qué necesita el cliente.

Todo lo que necesitas para transferencias perfectas

Seis capacidades core que hacen que cada transferencia sea fluida, rápida y sin pérdida de información.

🔄

Transferencia Fría

Redirección instantánea de llamadas sin intervención del agente receptor. Configuración de reglas de enrutamiento por hora, habilidad y carga de trabajo. Ideal para volúmenes altos donde la velocidad es crítica.

🤝

Transferencia Tibia

El agente receptor escucha la llamada antes de anunciarse. La IA genera un resumen estructurado del problema detectado, el historial del cliente y las acciones ya tomadas durante la conversación.

🧠

Agentic Transfer with Context

Transfer between AI agents with complete transfer of conversational state: intent, entities detected, steps completed, data collected and recommended next steps.

📡

Availability Detection

Consult the availability status of each agent or queue in real time before starting the transfer. If the target agent is not available, the system automatically searches for the best alternative.

📋

Automatic Context Summarization

AI generates a structured briefing in less than 200 ms with: reason for the call, customer data, sentiment detected, previous actions and recommendations for successful resolution.

🏗️

Multi-level Routing

Define complex routing rules with multiple criteria: language, account type, urgency, required skills, schedule and maximum waiting time. Up to 10 configurable escalation levels.

How the transfer works step by step

Four automatic steps that occur in less than 800 ms to ensure a seamless experience.

1

Need Detection

AI analyzes in real time the customer's intent, the level of complexity of the request, and whether the current agent can resolve it. Identify signs of escalation: frustration, supervisor requests, complex technical issues, or cases outside the bot's scope.

2

Availability Assessment

The routing engine queries the availability of qualified agents in real time. Evaluate: estimated wait time, required skills, history with the client and current load. If the preferred agent is not available, propose alternatives ordered by suitability.

3

Context Preparation

AI consolidates all conversation information into a structured briefing: executive summary, customer data, problem detected, actions taken, sentiment and tone, and recommended steps. This context packet is transmitted to the receiving agent before the call is connected.

4

Seamless Transfer

The call is transferred with all context. The receiving agent can review the briefing while the client is entertained with music or a personalized message. By connecting, the agent already knows exactly what the customer needs, completely eliminating the need for the customer to repeat their problem.

94%
Resolution by AI
without scaling to human
<800ms
Average transfer latency
40%
Reduced customer waiting time
0%
Context Loss
on transfers

Smart transfers in every industry

How different sectors take advantage of agentic transfer to transform their customer service.

🖥️ Escalated Technical Support

Top-level AI agent takes the call, diagnoses the problem, attempts basic resolution and when it detects that a specialist is required, it transfers with the complete diagnosis: analyzed logs, steps already attempted, type of error and level of urgency. The specialized technician enters directly into the exact point of the problem without wasting time in re-diagnosis.

✓ 55% reduction in technical resolution time

💼 Collection with Negotiation

The IA agent carries out the initial collection management, identifies the debtor's payment capacity, explores payment options and when it detects a need for special negotiation, transfers to the agent human with full profile: amount owed, initial customer proposal, payment history and authorized trading margin.

✓ 38% increase in portfolio recovery

🏥 Specialized Medical Care

In healthcare settings, the AI agent collects symptoms, verifies health insurance and pre-classifies the urgency. When transferring to doctor or nurse, send: reported symptoms, current medications, known allergies, level of urgency assessed and date of last contact. The health professional Come to the conversation fully informed.

✓ 70% less triage time in non-critical emergencies

🏢 Enterprise Sales

The bot qualifies the prospect during the first conversation: company size, budget, specific needs, decision timeline and contact authority level. When transferring to the enterprise sales executive, deliver the complete BANT qualification profile, objections already detected and the most relevant success stories for that profile.

✓ 45% more conversions on qualified demos

Infrastructure designed for speed and reliability

Each component of the transfer system is optimized to ensure that no call is lost and no context is degraded during the handover between agents.

P99 Latency Less than 800 ms for 99% of transfers, even during peak loads.
Supported protocols SIP, WebRTC, PSTN. Compatible with Twilio, Amazon Connect, Genesys, NICE.
Structured context Payload JSON with up to 50 context fields transmitted via webhook or REST API.
Automatic Fallback If the target agent is not available, the system automatically redirects in <100 ms.
Continuous recording Call recording is not interrupted during the transfer. Complete audit.
SLA guaranteed 99.9% uptime with automatic compensation in case of non-compliance with the service level.
Transfer architecture
🔵 Detection layer
NLP · Sentiment analysis · Intent classification · Complexity scoring
🟡 Routing Engine
Real-time availability · Skill matching · Load balancing · SLA priority
🟢 Context Generator
Summary LLM · Entity extraction · Action history · Next-step recommendation
⚡ Delivery <800ms
SIP transfer · Webhook payload · CRM update · Recording continuity
Total latency:
≤ 800 ms

Everything you need to know about transfers

What is the difference between cold, warm and agentic transfer?

+
The cold transfer directly connects the customer to the next agent without prior preparation: it is the fastest but the receiving agent has no context. Warm transfer allows the receiving agent to listen to the conversation or receive a summary before announcing to the customer, eliminating the need for the customer to repeat their problem. Agentic transfer goes further: AI generates a complete briefing with sentiment analysis, history, detected entities and recommended steps, allowing the receiving agent to arrive at the conversation fully prepared to resolve the problem in the shortest time possible.

How do you guarantee 800 ms latency in the transfer?

+
The 800 ms is achieved through three key optimizations: (1) intent analysis and handoff need detection run continuously during the call, not at decision time; (2) the context summary is generated in parallel with the availability evaluation; and (3) SIP signaling infrastructure is deployed in multiple regions with sub-50 ms latency between nodes. The result is that when the agent presses "transfer", 90% of the work is already done. We guarantee P99 latency of 800 ms with contractual SLA.

Does it work with my current call center system (Genesys, NICE, Avaya)?

+
Yes. NetVoice integrates with all major contact center systems using standard SIP protocols and REST APIs. We support Genesys Cloud and On-Premise, NICE CXone, Avaya, Amazon Connect, Twilio Flex, Five9, Zendesk Talk, and Salesforce Service Cloud Voice. The typical integration takes 3-7 business days and requires no modifications to your existing infrastructure. The transfer context is delivered via webhook or directly to the receiving agent's CRM through our native connectors.

What information exactly does the context packet sent to the receiving agent include?

+
The context package includes: (1) Customer Data: name, ID, segment, previous contact history; (2) Conversation summary: reason for call, problem detected, information already collected; (3) Emotional analysis: feeling (positive/neutral/negative/frustrated), tone of voice, level of urgency; (4) Actions taken: consultations made, solutions attempted, commitments made; (5) Entities detected: dates, account numbers, products mentioned, amounts; (6) Recommendations: suggested next steps, relevant resources, authorization required. All in structured JSON format and as a readable summary in natural language.

Ready to scale your voice operations?

Deploy intelligent voice agents in less than 30 days. No migrations, no downtime, no friction.