This startup says AI agents suffer from 'amnesia.' It raised $35 million to solve that with this pitch deck.

Graphon Labs, a San Francisco-based startup, has secured $35 million in funding to develop AI technology aimed at solving the problem of "amnesia" in AI agents. The company's product creates a connected "fabric" of information, mapping the relationships and changes among various data types such as documents, videos, and images. This differs from common AI search systems, which treat data as isolated snippets that are difficult to retrieve, resulting in lost context during searches.

Graphon's founder and CEO, Arbaaz Khan, compared traditional AI search systems to tearing pages out of a book, crumpling them up, and tossing them into a bin. He explained that when searching for answers, users are often left with a collection of disjointed pieces, as opposed to a cohesive "fabric" of information.

The company targets two primary customer groups: individual developers and businesses with vast amounts of technical and operational data, including industrial firms, financial institutions, and healthcare providers. Graphon's software is compatible with any off-the-shelf AI agents, allowing users to input large amounts of data without restriction.

Graphon Labs has already gained adoption from notable clients, including the San Francisco 49ers and HP. The company's $35 million Series A funding round was led by RRE Ventures, with support from Energy Impact Partners, Hitachi Ventures, Gaia VC, and HP. Earlier this year, Graphon closed a $8 million seed round.

The primary use of the funding will go towards hiring engineers, researchers, and sales personnel. Graphon plans to focus on enterprise customers before exploring personal devices and consumer products. The pitch deck used to secure the new funding can be accessed via the provided link.