This website uses cookies

Read our Privacy policy and Terms of use for more information.

A futuristic humanoid AI robot examines a large, glowing blue holographic display representing an AI agent's long-term memory. The interconnected digital interface features documents, photographs, videos, calendar entries, business charts, meeting images, notes, and code, all linked to a central layered data structure. The visualization illustrates how AI agents can retrieve, connect, and retain information across multiple formats over time.

Conceptual illustration of an AI agent accessing interconnected digital information, highlighting how persistent memory technology could help AI systems retain and retrieve knowledge over time. AI-generated image via ChatGPT (OpenAI)

Key Takeaways: Graphon Labs' $35 Million Series A and AI Memory Technology

Graphon Labs is an AI infrastructure startup that helps AI agents remember, connect, and build on information over time, across documents, videos, images, and audio.

  • Graphon Labs raised a $35 million Series A led by RRE Ventures, bringing total funding to $43.3 million.

  • Its software gives AI agents persistent memory by mapping how pieces of information connect, where they came from, and when they changed.

  • The round closed less than six months after an $8.3 million seed, a sign of how quickly investors are moving on AI agent infrastructure.

  • Graphon says the San Francisco 49ers, HP, and Hitachi use its technology, though some are still in early testing.

Today's AI agents can draft code, dig through company databases, and work through complicated, multistep jobs. What many of them still can't do well is remember.

Ask an agent about something buried in last quarter's files, or how a project changed over six months, and it can lose the thread. That gap is what Graphon Labs is trying to close.

The San Francisco startup announced on November 10th that it raised a $35 million Series A led by RRE Ventures, with participation from HP, Energy Impact Partners, Hitachi Ventures, and Gaia VC. The round brings its total funding to $43.3 million. Business Insider first reported the raise.

Why AI Agents Forget

Most AI tools handle large amounts of information in one of two ways. They either stuff as much as they can into the model's short-term working space, known as the context window, or they chop documents into small chunks and pull back the pieces that look most relevant, a method called retrieval-augmented generation, or RAG.

Both approaches work for simple questions. But as data piles up and changes, they can lose the bigger picture: how facts relate to each other, which came first, and what has since been updated.

CEO Arbaaz Khan compared common AI search to ripping the pages out of a book and then digging through the scraps for an answer, Business Insider reported. Each piece is there, but the story connecting them is gone.

How Graphon's AI Memory Technology Works

Graphon takes its name from graphons, a branch of mathematics for describing extremely large networks. Khan and cofounder Clark Zhang worked with the concept during their PhDs at the University of Pennsylvania, where Khan used it to coordinate swarms of robots.

Instead of treating files as isolated snippets, Graphon builds a map of the connections across a company's documents, videos, images, and audio, including where each piece of information came from and when it changed. A compressed version of that map then guides the AI to the right place when it searches.

Importantly, it isn't a new AI model. Companies plug their existing tools into it. "You can basically point your Claude or GPT or whatever to Graphon," Khan told Business Insider.

Graphon says its system scored 93.9% on LoCoMo, a test of memory across long conversations, while using 58% fewer tokens than feeding the entire conversation to the model. On ATM-Bench, a University of Cambridge benchmark spanning video, audio, images, and text, it says it scored 81.7%, about 5 points ahead of the next system on the public leaderboard.

Who's Using Graphon So Far

Graphon says the San Francisco 49ers, Hitachi, and other large companies use its software.

Some of that work is still early. A 49ers technology executive said the team is exploring uses such as spotting new sponsorship opportunities, and Hitachi Rail described its evaluation as promising for its industrial operations.

Why Investors Are Betting on AI Memory

The speed of the raise stands out. Graphon came out of stealth in May with its $8.3 million seed round and landed a Series A roughly six months later, with a team of just 20 people.

The investor list is telling, too. HP and Hitachi are both backers and users, a common pattern when large companies want early access to technology they may build into their own products.

RRE general partner Raju Rishi said the firm sees Graphon becoming "the source of truth beneath every agent."

The bet comes with competition. The biggest AI companies are adding memory to their own assistants, with Anthropic and Apple both expanding memory features in Claude and Siri this year. Graphon's pitch is that businesses need a memory layer that works across any model and any type of data, not one tied to a single company's chatbot.

Graphon frames its work as a step toward AGI. For most businesses, the near-term promise is more practical: AI tools that don't forget what they already know.

What's Next for Graphon Labs

Graphon plans to spend most of the money hiring engineers and researchers, along with some sales staff. It offers three products: Graphon Enterprise, a developer platform, and on-device memory for phones, laptops, cameras, and wearables.

Khan told Business Insider the company will stay focused on enterprise customers before pushing further into personal devices and consumer products.

Sources:

About Natalie Bartels

Natalie Bartels is a contributing writer at AiNews.com covering AI startups, venture capital, and emerging technologies. She is also the founder of Monstera Communication, a communications firm for AI and frontier technology companies.

Over the past decade, Natalie has led communications strategy inside top tech agencies and tier-one venture funds, overseeing more than 100 announcements for companies from seed to IPO.

Natalie has worked with founders across physical AI, AI models, science, and deep tech. Her writing explores how AI companies secure funding and how frontier AI technologies move from research labs to commercial markets.

Connect with Natalie on LinkedIn

Email her at [email protected].

Learn more about Monstera Communications