New York-based startup Cake is emerging from stealth mode today, backed by Google's AI-focused venture fund, Gradient Ventures, with a mission to simplify the integration and security of open-source AI infrastructure for businesses. Founded in 2022 by Misha Herscu and Skyler Thomas, Cake has already raised $13 million in funding, including a recent $10 million seed round led by Gradient Ventures.

Cake's platform integrates and secures over 100 open-source components, including data source adapters, data ingestion tools, data labeling platforms, vector and graph databases, generative AI APIs, and more. This comprehensive approach allows businesses to bypass the complexity of building and maintaining their own AI stacks, reducing engineering overheads and accelerating AI adoption.

The company's founders, Herscu and Thomas, bring extensive experience in AI and enterprise technology. Herscu previously founded McCoy Medical Technologies, an AI company focused on machine learning infrastructure for radiology, which was acquired by IT vendor TeraRecon in 2017. Thomas, Cake's CTO, has a background in enterprise technology, having worked at IBM and Hewlett Packard Enterprise, where he was a distinguished engineer and director of strategy.

Cake's solution addresses the "big picture problem" faced by businesses when integrating open-source AI components. According to Herscu, the challenge lies not in individual components but in integrating them reliably and making them production-ready. Cake's platform provides a curated selection of open-source projects, assembled and managed to ensure seamless integration and security.

The company's platform has already attracted customers like AI bioscience startup Altis Labs and data intelligence insurtech Ping. Cake's sweet spot lies in supporting businesses that require custom AI solutions beyond what off-the-shelf products can offer. This includes use cases like natural-language query processing, medical image analysis, and recommendation engine upgrades.

Cake's CTO, Thomas, draws parallels between the company's mission and the early days of Linux, where thousands of open-source packages needed integration and security. Red Hat, acquired by IBM for $34 billion, made Linux safe for the enterprise, and Cake aims to do the same for AI.

While Cake's platform currently requires on-premise deployment, the company plans to introduce a hosted version in the future. This will cater to organizations with lower compliance obligations, although many businesses will continue to require on-premise deployment due to data privacy stipulations.

The $10 million seed round, which closed in April, demonstrates the company's traction and the confidence of its investors, including Gradient Ventures, Primary Venture Partners, Alumni Ventures, Friends & Family Capital, Correlation Ventures, and Firestreak Ventures. Cake is already looking toward its next financing round, with plans to raise again in mid-2025.

With its innovative approach to simplifying open-source AI infrastructure, Cake is poised to make a significant impact in the enterprise technology landscape. As the company continues to grow and expand its platform, it will be interesting to see how its solution resonates with businesses seeking to harness the power of AI.