Supermemory raises $3 million with the best memory engine for LLMs
Today, I am excited to announce our first funding round to accelerate our mission of building an interoperable, scalable and reliable memory for LLMs and agents.
Memory is one of the hardest challenges in AI right now. We have really intelligent models (Claude, GPT-5, etc) and tools (Cursor, and many others), and I genuinely believe that the next hard frontier will be personalization. And that's the problem we're solving at supermemory - self-learning context about your users that is interoperable any model.
It all started as a consumer app built over the summer last year - to help humans organize the knowledge they collect on various apps. At the time, Supermemory was not trying to be a business. It was an open-source "second brain" app that users loved. We quickly reached 50k+ users, millions of items saved by these users, and 10,000 stars on Github, and it was one of the fastest growing OSS projects in 2024. Very quickly, this side-project turned into a real business.
At scale, the consumer app ran into many issues - and to our surprise the infrastructure for "Memory" for LLMs like this simply didn't exist. I had some experience in infrastructure and started sharing more details on how we were building the infrastructure behind the consumer app ourselves.
At the time, I was working on AI infrastructure at Cloudflare, where we filed a patent to make agents faster. I worked at startups working on memory, made multiple consumer-led apps, and much more. Doing this, I realized the real problems with memory - it's not just a search problem, it's about really understanding the users and making their experience magical by contextualizing the LLMs they talk to.
While many loved the consumer app, and the project won the buildspace grant, among many others, a lot of interest came from companies wanting this infrastructure for their products. Many were ready to pay right away, and many offered contractual or consultancy work for help with setting up the open-source project, and that's when we decided to open up the infrastructure for everyone to use.
We built our own vector database, content parser and extractor, and a lot of infrastructure components were built from scratch, purpose-driven for being a flexible, scalable memory layer that works like the human brain.
After all this work, Supermemory managed to top every benchmark and extremely scalable while providing some of the best latencies in the market.
Today, we have the best, scalable and fastest memory engine.
Ever since then, our customers have been loving supermemory in every way. Today, some send us many billions of tokens every week, and our product is loved by enterprise customers like Cluely, Composio etc. as well as open source projects like Scira AI.
Today, I am excited to announce the next chapter of supermemory - We have raised $3 million dollars led by Susa Ventures (Shaheer), Browder Capital, SF1.vc, with participation from top angels like Jeff Dean (Chief Research Scientist at DeepMind), Theo Browne, Dane Knecht (CTO Cloudflare), David Cramer, Julian Weisser at Solo founders program, and others leading in the space of infrastructure and AI.
We are hiring across engineering, research and product roles. Join us in the journey of creating the best memory engine on 🌎