Voice-First CX Platform Omilia Raises $67M Series B

Enterprise adoption of customer-service AI is moving beyond chatbots. AI agents are increasingly expected to handle live calls, authenticate customers and complete transactions across regulated workflows.

Omilia, a voice-first enterprise CX company with more than two decades of operating history, has announced a $67 million Series B round led by Expedition Growth Capital. Completion remains subject to customary conditions.

Two Decades in Enterprise Voice AI

Omilia was founded in 2002 by Dimitris Vassos, who began his career at IBM UK and worked on voice-technology deployments across more than 70 countries.

John Nikolaidis, Omilia’s founding partner and chief commercial officer, brings more than 20 years of international business-development experience across mobile and IT services.

The company grew profitably without institutional capital for nearly two decades. It raised its first external financing in 2020, when Grafton Capital led a $20 million growth round. Expedition founder Oliver Thomas led that investment while at Grafton and has remained involved as a board director.

Omilia says live annual recurring revenue has increased more than tenfold since its Series A and now exceeds $60 million. It did not disclose a valuation for the new round.

Built for High-Volume Voice Operations

Omilia’s Self-Learning Agentic CX platform analyzes real customer interactions, identifies successful practices and failures, and proposes improvements to customer-service agents. Changes can be tested through simulation and introduced with human oversight.

Unlike vendors that primarily connect frontier language models with third-party speech and workflow services, Omilia says it owns the core stack spanning speech, reasoning, routing, workflow execution, analytics and continuous learning.

That architecture also supports a pricing model without passing token costs directly to customers. The pitch is particularly relevant to banks, insurers and healthcare providers that require predictable costs, data control and auditability.

Omilia says one major US bank uses its platform to handle more than one million calls a day. A multinational insurer operates an agent-assist deployment covering more than 600,000 daily calls. The company also claims it can support more than 50,000 concurrent voice interactions for a single customer with sub-second latency.

Customers include Capital One, Discover, RBC, Taco Bell, the UK Department for Work and Pensions and PSEG. Its compliance portfolio includes FedRAMP, PCI-DSS, SOC 2, HIPAA and GDPR.

From Contact Centers to Drive-Thrus

Omilia’s technology is also deployed across more than 1,000 Taco Bell drive-thrus in 38 US states. The system must process road noise, regional accents, order changes and store-specific menus and inventory in real time.

In July, the company launched Lexis, a generative text-to-speech model built directly into the Omilia platform. Lexis supports more than 25 locales and can create customized brand voices without routing audio through an external speech API.

The new capital will support expansion in North America and other international markets, including Omilia’s first US office in the second half of 2026.

Competing With Sierra and Decagon

The customer-service AI market is attracting increasingly large rounds.

Sierra combines multiple frontier models with its own orchestration layer to create enterprise customer-service agents. The company raised $950 million in May at a $15.8 billion valuation and reported ARR above $150 million.

Decagon allows CX teams to define agent behavior through natural-language Agent Operating Procedures. It raised $250 million in January at a $4.5 billion valuation.

Parloa is pursuing an omnichannel platform spanning voice, chat, applications and the web. It raised $350 million in January at a $3 billion valuation. PolyAI is another close competitor with a strong focus on automated enterprise phone conversations.

Omilia’s distinction is its proprietary, voice-native stack and long record of production deployments in regulated industries. It is selling control, compliance, high concurrency and cost predictability rather than relying primarily on access to the latest frontier model.

Owning the stack also creates a challenge. Omilia must continue funding its models and infrastructure while keeping pace with rapid improvements from external model providers. Its $60 million ARR demonstrates meaningful commercial traction, but competitors are deploying substantially larger pools of capital into product development and US distribution.