Continua, a startup developing artificial intelligence assistants designed specifically for group conversations has secured $8 million in seed funding. The funding round was led by GV, with participation from Bessemer Venture Partners and notable angel investors including Jeff Dean, Spencer Kimball, Eoghan McCabe, Lauryn Motamedi.
The Continua platform addresses what the company identifies as a significant gap in current AI applications. While chatbots excel in one-on-one interactions, they remain largely absent from collaborative environments where multiple people communicate simultaneously.
From Google Glass to Group Intelligence
David Petrou founder and CEO of Continua AI, brings extensive experience in contextual computing from his tenure at Google, where he served as a Distinguished Software Engineer. His previous work included founding contributions to Google Goggles, the predecessor to Google Lens, and participation in the original Google Glass development team.
“I was a serial entrepreneur at Google: I came up with product ideas, prototyped them, evangelized them, built teams around them” David explained. His focus centered on what he terms “personal AI” – systems that understand and represent individual users through screen understanding and privacy-preserving machine learning.
The transition from corporate research to startup founder reflected David’s desire to operate under resource constraints and accelerated timelines that large technology companies cannot replicate, even within their internal innovation programs.
Social Context as Missing Element
The inspiration for Continua emerged from David’s recognition that existing AI systems lacked crucial social context. While previous projects like Google Glass could interpret external environments, group communications represent what he considers far more valuable contextual information.
Continua functions as what the company describes as a “quiet helper” integrated directly into existing group messaging platforms. Users can add the service to group chats on platforms like Discord or through phone numbers for text messaging, eliminating the need for separate applications.
The system has processed over 7 million messages during its initial release period, demonstrating what the company reports as strong user retention metrics.
Technical Architecture and Capabilities
The platform’s functionality extends beyond simple question-answering to include proactive assistance and memory capabilities. Continua can capture important information from conversations, generate and maintain collaborative documents, and provide private consultation through direct messaging without disturbing group dynamics.
According to David, the technical implementation required significant modifications to standard large language model architectures. “We sort of had to break the LLM’s brain to get the kind of social intelligence and etiquette that is needed” he noted.
The system employs multiple fine-tuned models rather than relying on a single large language model, with the company filing patents on its approach to group conversation AI.
Privacy and Integration Approach
Continua emphasizes privacy protection through data encryption both at rest and in transit, drawing on David’s team’s experience developing privacy policies and technologies at major technology companies.
The integration strategy focuses on meeting users within their existing communication workflows rather than requiring adoption of new platforms. This approach allows the AI to participate in group conversations as what David characterizes as a “trusted member” that can distinguish between relevant information and distractions.
Research and Development Focus
The company positions social AI as an emerging research area requiring continued breakthroughs. Unlike traditional AI applications designed for single-user interactions, group conversation AI must navigate complex social dynamics while adding value without disrupting natural communication flows.