Deepgram’s award-winning voice AI goes global with Dedicated and EU-hosted deployments 🌍

Article·AI Agents·Aug 4, 2025
5 min read

Vonage and Deepgram: Enabling Real-Time Translation for Contact Centers

Vonage and Deepgram have joined forces to bring mulitlingual voice agents to life. Together, we’ve developed a proof-of-concept that combines telephony, voice AI, and real-time translation into a single, seamless workflow.
5 min read
Jose Nicholas Francisco
By Jose Nicholas FranciscoMachine Learning Developer Advocate
Last Updated

Breaking Down Language Barriers: Real-Time Translator for Contact Centers

As contact centers scale across borders, language mismatches remain a persistent source of customer frustration and operational inefficiency. 

Picture a Spanish-speaking caller needing support from an English-only agent, or a French customer dialing into a Mandarin-speaking team—these everyday scenarios often end in dropped calls, miscommunication, and lost revenue. But what if agents could speak naturally in their own language, while customers instantly heard responses in theirs?

Vonage and Deepgram have joined forces to bring this vision to life. Together, we’ve developed a proof-of-concept that combines telephony, voice AI, and real-time translation into a single, seamless workflow. 

The result: an experience where every conversation—no matter the languages spoken—is clear, immediate, and human. Here's how it works—and why it has the potential to reshape global customer support.

The Challenge for Call Centers

For global contact centers, language remains one of the biggest operational hurdles. Traditionally, agents must be matched with callers who speak the same language—a requirement that narrows the talent pool, increases staffing costs, and complicates scheduling. 

For customers, the experience is just as frustrating: navigating multi-step IVRs or being rerouted between language-specific teams often leads to abandoned calls, dissatisfaction, and lost trust. 

Behind the scenes, maintaining separate queues, training materials, and workflows for each supported language adds significant complexity, making it harder to deliver fast, consistent support at scale.

Real-time Voice Translation: A Seamless Solution to the Language Barrier

By integrating best-in-class technologies, Vonage and Deepgram offer a solution that removes language as a barrier—without overhauling existing workflows:

First, Vonage orchestrates the call setup, routing, and audio streaming through its powerful Communications APIs. 

Meanwhile, Deepgram processes both the agent’s and caller’s audio streams in real time, delivering ultra-low-latency speech-to-text and text-to-speech on both sides of the conversation. 

This enables each participant to speak naturally in their own language, while hearing translated responses near-instantly—creating a fluid, bilingual experience that feels as effortless as a single-language call.

How Speech to Speech Translation Works

  • 📲 Call Initiation: A caller dials into the Vonage virtual number. Vonage routes the media to a lightweight orchestration layer.

  • 🔊 Live Transcription & Translation: Deepgram converts incoming speech (e.g., Spanish) to text in under 200 ms. Vonage translates that text into English and returns it within an additional 100–150 ms.

  • ▶️ Agent Response: The English text is fed back to Deepgram, which synthesizes it as speech for the original caller in Spanish. Simultaneously, the English agent sees real-time subtitles of the caller’s Spanish.

  • 💬 Seamless Two-Way Flow: Both parties speak naturally in their own languages, oblivious to the AI “interpreting” mid-call.

Key Benefits of Real-Time Voice Translation

This integrated approach unlocks significant advantages for contact centers worldwide. 

By removing language constraints, hiring teams can broaden their talent pools, recruiting fluent English speakers, bilinguals, or any language combination without the need to maintain separate queues. 

Real-time translation streamlines interactions, eliminating costly transfers and callbacks to boost First-Call Resolution (FCR) rates. 

Operational costs also decline, as the solution reduces the complexity of multilingual training and IVRs, allowing contact centers to scale language support on demand—without proportional increases in headcount. 

Most importantly, customers experience more authentic, effective conversations, leading to higher Net Promoter Scores (NPS) and stronger loyalty.

Get Involved
Ready to break language barriers in your contact center?

With this collaboration, every call becomes an opportunity—regardless of language.

Unlock language AI at scale with an API call.

Get conversational intelligence with transcription and understanding on the world's best speech AI platform.