Guava secures $12.5 million seed funding to scale its AI voice agent platform

# Guava Raises $12.5 Million in Seed Funding to Expand Enterprise Voice AI Technology

Artificial intelligence startup Guava has raised $12.5 million in seed funding as it looks to improve voice AI technology for businesses handling large volumes of customer conversations. The financing round was led by Acrew Capital, with participation from Flourish Ventures, WillowTree Ventures, Clocktower Ventures and existing pre-seed investor Commerce Ventures. The investment will support the company's efforts to develop its voice models, expand its engineering capabilities and bring its technology to more enterprise customers.

Guava's first publicly available model, called Daytona, was launched in September and is already being used by more than a dozen enterprise customers. According to the company, the system is powering tens of thousands of calls every day, indicating early demand for its approach to automated voice interactions. Guava aims to help businesses deploy AI agents that can manage customer conversations reliably, particularly in industries where accuracy, speed and consistency are essential.

The company builds on more than a decade of voice AI research conducted by Gridspace, a voice technology research lab. Gridspace trained its models using more than 10 billion minutes of real-world conversations from regulated call centers. This extensive training experience provides the technical foundation behind Guava's products and is intended to help its systems perform effectively in complex customer service environments.

Guava was co-founded by Omri Dahan, Anthony Scodary, Nico Benitez and Mike Piechna. Chief executive officer Dahan said the company was established after recognizing that many businesses remained dissatisfied with the voice AI products they had tested. According to him, existing solutions often perform well during demonstrations but struggle when deployed in real-world situations involving high volumes of calls and sensitive customer interactions.

Dahan said Guava wants to give enterprises voice agents they can trust to handle important conversations at scale. He also explained that the latest funding would help the company build its go-to-market team and reach more businesses and business process outsourcing (BPO) providers interested in moving voice AI projects beyond experimental pilot programs into full production.

Acrew Capital founding partner Lauren Kolodny said the investment was driven by the strength of Guava's technical team and Dahan's business experience. She noted that the founders had been developing voice AI technology for more than a decade, well before the current surge in interest surrounding artificial intelligence. The combination of long-term research expertise and leadership focused on identifying market opportunities made the company an attractive investment, she said.

Guava is taking a different approach from many existing voice AI platforms, which typically combine products from several technology providers. Such systems often depend on separate vendors for speech recognition, language processing and voice generation, while developers control much of the agent's behavior through prompts. Guava instead develops and operates its own speech recognition systems, language models and speech synthesis technology as a unified platform.

By managing the entire technology stack, the company says it can control the voice experience from the moment a customer begins speaking until the conversation ends. This approach is intended to reduce delays, improve response quality and make interactions feel more natural. It also gives the company greater control over how its different AI components work together, rather than depending on external providers and their individual limitations.

Co-founder and co-head of engineering Anthony Scodary said relying on models from multiple vendors can introduce additional latency, technical failures and dependence on outside companies' development schedules. Guava trains and operates its own models, allowing its engineers to optimize the complete system around the goal of delivering natural conversations that successfully resolve customer requests.

The company is also developing a software development kit (SDK) designed to make voice AI easier for developers to integrate into their applications. Instead of requiring engineering teams to become specialists in managing voice agents and writing complex prompts, Guava's platform allows them to build interactions using business rules and code. This approach is intended to give companies greater control over how automated agents respond to customers and complete specific tasks.

The new funding will also support Guava's plans to expand its technology into additional applications, including edge devices and on-premises installations. The company is also working toward full-duplex voice interaction, which would allow an AI agent to listen and speak simultaneously, more closely resembling the way people communicate in natural conversations.

Guava's funding comes as businesses increasingly explore AI-powered customer service tools to reduce operational workloads and improve the availability of support services. However, deploying voice agents in real-world environments remains challenging, particularly when conversations involve interruptions, complex requests or sensitive information. Companies need systems that can understand customers accurately and respond consistently without creating frustrating experiences.

With its combination of long-term research, proprietary models and a developer-focused platform, Guava is positioning itself as a competitor in the growing enterprise voice AI market. The $12.5 million investment will help the startup move toward broader commercial adoption while advancing its technology for more demanding business applications. Its ability to deliver reliable performance at scale will be an important factor in determining whether it can become a major platform for enterprise voice automation.