Fintech company Revolut is a financial services company with more than 70 million customers. Revolut put ElevenLabs Agents on the front line of its UK and European voice support, and said that in the initial rollout it cut ticket resolution time by more than 8 times compared with its previous support. ElevenLabs announced the case on its official blog and on X (January 28, 23:59 KST) on January 28, 2026. It is a case of connecting an AI voice agent to real customer support in an environment like finance, where security and trust matter.
Source: X @ElevenLabs (2026-01-28 23:59 KST)
What the Revolut AI support agent handles
ElevenLabs Agents is an AI support agent platform that understands and responds to real-time voice conversation. Revolut placed it at the first step of voice support in the UK and Europe, bringing more than 4 million customers into scope. Support happens over phone voice calls.
During a call, the agent detects the language the customer is using and switches in real time when needed. It answers FAQs tied to account context and uses live customer data to give tailored responses. Where possible, it handles disputes and chargebacks from start to finish. ElevenLabs says it operates in more than 31 languages, responds to customers instantly, and provides 24-hour support. You can read more about the product concept in the AI agent glossary entry.
Revolut keeps control of its own orchestration and business logic and connected it to its existing chat and digital AI stack through an API. The structure has ElevenLabs Agents handle the voice interaction, while the workflow and permission boundaries stay in Revolut's systems.
Resolution within 5 minutes, call success rate above 99.7%
These are the results from the first rollout that ElevenLabs published. The company said that when comparing averages, the human agents' handling time included queue time, and that it did not adjust for inquiry difficulty.
| Metric | Official announcement | Interpretation |
|---|---|---|
| Ticket resolution time | Within 5 minutes | More than 8 times faster than previous agent handling |
| Call success rate | Above 99.7% | Based on calls the agent handled |
| Languages | 31+ | Real-time language detection and switching |
| Coverage | Over 4 million | Customer coverage of the first UK and Europe rollout |
Chose ElevenLabs after comparing several vendors
Revolut first built its own prototype. But connecting speech recognition (STT), a language model (LLM), and speech synthesis (TTS) into real-time conversation, and building production-grade turn-taking, testing, monitoring, PCI compliance, and zero-retention data controls itself, was a heavy burden at platform scale. So it said it compared the major vendors directly and then chose ElevenLabs.
Ines Azevedo, Revolut's Tech Product Owner (AI), explained that during evaluation some vendors fell short on flexibility, some on reliability targets, and others on the voice quality needed. According to her, ElevenLabs delivered voice quality and low latency while letting Revolut keep control of orchestration, and its PCI compliance and zero-retention option also matched security needs.
Julia Ponomareva, Director of CX and AI Products, explained that while delivering complex financial services in nearly 40 markets and multiple languages, it needed a partner that could protect both speed and quality. ElevenLabs says it also provided Forward Deployed Engineers support during the rollout.
ElevenLabs co-founder and CEO Mati Staniszewski also announced the adoption on X.
Source: X @mati (2026-01-29 01:04 KST)
How to adopt ElevenLabs Agents, and pricing
ElevenLabs' product docs now call the product "ElevenAgents." You can configure agents through the dashboard, the API, the Agents CLI, or an MCP server and deploy them to phone, web, and mobile. You can find the setup and testing steps in the product overview and the quickstart docs, and the quickstart walks through building a first agent in about 5 minutes.
As of September 28, 2026, the official pricing page lists monthly self-service plans as Free with 15 minutes, Starter at $6 for 75 minutes, Creator at $22 for 275 minutes ($11 the first month), Pro at $99 for 1,238 minutes, Scale at $299 for 3,738 minutes, and Business at $990 for 12,375 minutes. Extra calls are $0.08 per minute, calls over the concurrent call limit are $0.16 per minute, and text messages are $0.003 each. LLM costs are separate. Pricing and included usage can change, so check the official pricing page before adopting.
Enterprise has custom pricing, and you can talk with the sales team on the Enterprise contact page. Organizations that need large-scale concurrent calls, custom contracts, and enterprise support can discuss adoption terms through this route.