# Graphon announces $35M Series A for AI agent memory

> Original title: Memory is one of the biggest open problems for AI agents right now.

- Source: Elvis Saravia
- Published: 2026-11-10T17:17:16.000Z
- Bad Influence: https://badinfluence.news/items/fku4d3ytfjj87v20gq2pa03ln
- Original: https://x.com/omarsar0/status/2120204152479400099

## Summary

Graphon announces a $35 million Series A, six months after its seed round, to build AI agent memory. Co-founder Elvis Saravia says the company builds one graph across a company's documents, emails, images, and videos. Per the post, Graphon scores 94% on LoCoMo using 7.9K tokens of context and ranks first on ATM-Bench.

## Why it matters

The source gives the funding amount, round, and benchmark results, showing how a memory startup for AI agents positions its product against rivals.

## Post

Memory is one of the biggest open problems for AI agents right now.

Graphon builds one graph across a company's documents, emails, images, and videos without a hand-made schema, so new data does not break it.

On LoCoMo, it reaches 94% accuracy with only 7.9K tokens of context, higher than giving a frontier model the full history. It is also #1 on ATM-Bench, which tests years of real emails, photos, and videos.

![](https://badinfluence.news/api/img-proxy?u=https%3A%2F%2Fpbs.twimg.com%2Fmedia%2FHWx6Qd5WEAAKhmP.jpg&mode=thumb&exp=1794528000&sig=8c5b309fa56e3895)

## Quoted post @graphonfounder

> After years working on robot memory at @NASAJPL, @GoogleDeepMind , @Meta  and @nvidia, @clarkzhango and I kept asking: why does AI still forget?
> 
> Today, with @deepakmishraVC, we’re announcing @GraphonAI and our $35M Series A just 6 months after our seed.
> 
> • #1 on ATM-Bench, ATM-Bench Hard
> • 94% on LoCoMo
> • 84% on DolphinBench, 10x cheaper
> • A new cost-accuracy frontier across 7 benchmarks
> 
> https://www.businessinsider.com/graphon-pitch-deck-funding-ai-agents-amnesia-2026-10

https://x.com/graphonfounder/status/2120201264932003990