Persistent memory is the Achilles heel of AI. Engramme’s Large Memory Models (LMMs) empower every app with persistent memory. Google solved search. OpenAI solved language. Engramme solved memory. Join beta:

Very interesting direction. Proactive memory recall feels like a fundamental primitive of modern AI agents. I shouldn't have to search for context that exists already. The brain doesn't work like that, and AI agents shouldn't either. Great launch!

Agreed. Also, the problem will be exacerbated as agents start to live longer lives. Compare current lifespans for agents versus humans and therefore their memory retrieval needs.

Not open source Our agents already have this Good luck

Trust matters here. Memory tied to real events, not generated text, is the only way enterprise will adopt this.

Agreed! Unlike LLMs, Large Memory Models do not hallucinate, and allow perfect recall.

Exciting stuff! Memory is such a game changer for AI. Can't wait to see how this opens up new possibilities for apps!

Read more about our thoughts on human memory augmentation here:

Or just use mine for free, completely open sourced

This is brilliant! "Google solved search. OpenAI solved language. Engramme solved memory." What an impactful way to frame the challenge and the solution.

Just signed up for the beta

What's the latency on memory retrieval? Real-time recall is a high bar.

Large Memory Models enable sub 1s latency!

We've been at it with OpenHuman - 10M tokens of context, 4000 tok/sec, runs locally on your machine. $0 raised, fully open source. Different bet (local-first, you own the memory) but same conviction. Live now:

The pitch is brilliant, @gkreiman 👏 super curious how the API handles privacy and data fencing across different apps :)

Maybe you should read this about persistent memory

This idea of memory showing up without searching is actually what AI has been missing. 🫡

Super interesting. Been waiting for something like this!

Samsung, Superhuman, Dropbox, and Microsoft are already testing integrations. The battle for the memory layer is being decided right now

I need this!

Signed up for beta 👌

This is going to be revolutionary for sure! 🚀

If you're curious, here are our thoughts on how humanity will change with perfect, infinite memory: Read more about our thoughts on human memory augmentation here:

It makes sense that someone would do it. Engramme appears to be the first with a new memory architecture.

Going to sign up for the beta

tbh video felt like a breath of fresh air in the sea of fancy launch video slop lol congrats on the launch :)

Is the memory beta open to individual developers or enterprise only?

We have an ongoing private beta! Please sign up here:

What I see is No arxiv paper from Kreiman or Madan on “Large Memory Models” specifically No benchmark numbers (no LOCOMO, no BABILong scores) No latency / recall@k / cost numbers Just hype. Interesting founder raising money. Thats it

Persistent memory is definitely one of the biggest missing pieces in AI today interesting to see Engramme tackling it head-on.

In 5 years every app will have a memory layer. Engramme is building the picks and shovels.

AI finally getting a memory layer—this changes everything.

“Solved” is ballsy

Proactive recall across apps sounds like a massive unlock for real productivity.

Finally!

This is going to be revolutionary for sure!

This is actually interesting.

This is going to be amazing

Count me in. The memory pitch alone is brilliant.

What makes @EngrammeHQ the trust layer I shoud let allow to ‘handle all of my ideas’? You are not the first to pitch “remember everything” I’ve tried it all Early adopters of: @LimitlessAI Panicked as it SOLD OUT to Meta. so what memory problem do you solve?

I signed up for beta

Memory singularity. Best framing I've heard in months.