We're partnering with Deutsche Telekom to integrate AI voice technology across its business. Europe's largest telecommunications company is bringing AI into the contact center, to its consumer app, and into the network itself. Its AI call assistant runs on ElevenLabs, embedded in the network so it works on any mobile device that can make a call. That brings customers: - Live assistance during calls - Real-time translation - Call transcription and summaries Deutsche Telekom is also using ElevenLabs in their customer service operations. The voice quality keeps customers on the line with the AI agent, giving the agents the opportunity to resolve their questions end to end. What began as a set of app features in early 2025 has grown into an integration at the core of Deutsche Telekom's services - and our partnership continues to expand.
Really impressive this makes voice AI feel genuinely useful, not just like another app. Bringing live translation and assistance directly into the call, even on existing devices, is a huge step. Curious how you’re balancing real-time latency with privacy at network scale
This is a strong example of AI moving from standalone applications into core communications infrastructure. Embedding voice intelligence directly into the network can make translation, transcription and assistance effectively ubiquitous, while contact-centre deployment shows how conversational AI can create value when it improves resolution quality rather than merely automating interactions.
Network-level integration is a different category than an app feature. Embedding AI voice into the network itself means it works on any device that can make a call - no app download, no software update, no user behavior change required. That's how you get telco-scale adoption. Not by convincing users to try something new, but by putting it in the path of what they already do. Having spent years on the enterprise side of telecom, I know how difficult Deutsche Telekom-level integrations are to close. When Europe's largest telco goes from app features in early 2025 to core infrastructure integration by mid-2025, that's not a pilot - that's a strategic commitment. The voice quality keeping customers on the line is the right metric. Resolution rate follows from there.
The part worth underlining is the shift from app features to network-level integration. Once the AI call assistant can work across devices that already make calls, distribution becomes part of the product experience rather than another workflow users have to adopt. That makes the use cases here especially interesting. Which do you expect to drive adoption first: real-time translation, live assistance, or call summaries?
The contact center is the right proving ground precisely because the calls are high-stakes, identity, billing, disputes, not just easy queries. Voice quality is table stakes now, what decides adoption at telecom scale is whether the system handles the messy, regulated turn reliably and hands off cleanly when it shouldn't be guessing. Putting it into the network itself is the ambitious part, that's where latency and trust stop being product features and become infrastructure.
Telecom-grade voice AI is where it gets real. We see the same pattern with property teams, production reliability beats demo polish every time. Curious how they're handling fallback to human agents at scale.
This is a great example of AI moving from “cool feature” to infrastructure-level adoption. Real-time translation, transcription, and voice assistance become much more powerful when they’re built directly into the network rather than sitting behind another app or workflow. At SynapseInfluence, we work at the intersection of AI, tech, SaaS, and creator-led storytelling, so it’s exciting to see voice AI becoming something people can experience naturally without changing how they communicate. The progression from app features to core telecom infrastructure is the real story here. 🚀
Category creation is one of the most difficult achievements in tech innovation. Embedding real-time AI voice and translation directly into the carrier network — rather than keeping it gated inside standalone apps — is a massive milestone for accessibility. Massive congratulations to both teams.
embedding this directly at the network level is the real game changer here... moving from basic conversational tools to actual autonomous action is exactly where the industry is heading...
Perhaps that's the real AI revolution — not when everyone starts using an AI app, but when billions of people use AI every day without even thinking about it.