Recall.ai's $38M Series B Funds Unified Conversational API

Recall.ai, a startup building foundational data infrastructure for conversational AI, has secured $38 million in a Series B funding round, bringing its valuation to $250 million. The round, led by Bessemer Venture Partners with new investments from Salesforce Ventures and HubSpot Ventures, validates the company’s strategic bet that developers will choose to buy, not build, the complex “picks and shovels” needed to tap into spoken conversations, according to a report from SiliconANGLE. This latest infusion of capital, part of a recent wave of AI developer tools funding news, will fuel Recall.ai’s expansion beyond online meetings to capture data from phone calls and in-person interactions, a key use for the new funding as noted by InfoTechLead. This move solidifies its role as a critical plumbing layer for a new generation of AI applications that can listen and understand. The company provides a unified API and SDK to help developers reliably capture and process conversational data from platforms like Zoom, Teams, and Google Meet.

Tapping the Untapped Conversational Goldmine

Recall.ai’s growth is predicated on a simple but powerful premise. In a recent interview, co-founder and CEO David Gu called conversational data “the world’s largest untapped dataset,” estimating there are “five times more words spoken at work in a single year than all the data on the internet.” With over 400 billion hours of work conversations happening globally each year, the vast majority of this data remains inaccessible to the AI models that need it to perform tasks like updating a CRM or drafting clinical notes.

Accessing this data is a formidable engineering challenge. It requires integrating with a fragmented ecosystem of meeting platforms, each with unique APIs, and managing computationally intensive audio and video processing. Gu notes that approximately 70% of workplace communication happens via phone or in-person, further complicating data capture. Recall.ai’s core value is solving this foundational problem, allowing developers to focus on application features rather than data ingestion infrastructure.

Digital Pipelines for Verbal Data

Recall.ai’s infrastructure stack is composed of two primary components designed to abstract away the complexity of data capture. The first is a unified API that provides a single point of access to recordings, transcripts, and metadata from numerous meeting platforms. This system is engineered for significant scale, processing over three terabytes of raw video per second at peak load and launching more than 8 million EC2 instances monthly. This level of conversational AI data API funding enables such resource-intensive operations, which allow developers to access transcripts within 10 seconds of a meeting’s end.

The second component, a Desktop Recording SDK, expands capture capabilities beyond meeting bots. By recording locally on a user’s device, the SDK offers universal compatibility with any desktop application, bypasses the need for a bot to join a call, and ensures reliable uploads even on poor networks. This approach supports rich data capture, including speaker identification and real-time video, without the obtrusive presence of a third-party bot.

Engineering Economics: Make vs. Source

The central thesis behind the Bessemer investment in Recall.ai is that outsourcing this complex infrastructure is a clear win for development teams. The company’s primary competition is not another API provider but the internal engineering resources of its potential customers. By offering a solution that reduces what could be a months-long project to just days, Recall.ai makes a compelling economic argument. Development teams using the platform report a 2-3x improvement in time-to-market.

This strategic positioning is reinforced by strong market validation from both investors and customers. The participation of Salesforce and HubSpot Ventures underscores the importance of conversational intelligence in the enterprise software ecosystem. With a customer list that includes HubSpot, DataDog, ClickUp, and Apollo.io, and a suite of compliance certifications like SOC2, HIPAA, ISO 27001, GDPR, and CCPA, Recall.ai has established the trust necessary for enterprise adoption.

Building the Backbone of Voice Intelligence

Recall.ai’s Series B funding is a significant indicator of the maturation of the AI stack. As large language models become commoditized, the unique, high-quality data needed to power them becomes the key differentiator. By focusing on the difficult engineering of data capture, Recall.ai is building an essential utility layer for any application that needs to understand human speech. The company’s roadmap, which reportedly includes a Mobile SDK for in-person meetings and VoIP integrations, points toward a future where all conversational contexts are accessible. As this data pipeline becomes a standard component, what new classes of AI-native applications will emerge when software can truly listen?