Jack Morris and a team of AI researchers have launched Engram out of stealth with $98 million to tackle how models handle memory and keep learning from new interactions instead of resetting or forgetting.
finally sharing what i've been up to! left phd end of 2025 and co-founded Engram.
there are a few startups in SF right making very different bets on the right way to train AI models. this is ours:
people want models that learn over time, remember details, adapt and interact like a person would
everyone gets a model. your model updates ~every minute. this is the world we're building. :)
http://x.com/i/article/2069463677733142528
The round pulled in General Catalyst, Sequoia, Kleiner Perkins and angels such as Andrej Karpathy, signaling strong belief that the group's prior work on retrieval, fine-tuning and state-space models can translate into practical memory layers.
Early traction centers on partners like Microsoft, Notion and Harvey for agent and workflow memory tools, while claims of minute-by-minute per-user model updates remain unconfirmed outside social posts.
Many users congratulated Jack Morris on co-founding Engram for personal AI models that update every minute, while a few dismissed the idea as just another limited fix for LLMs.
No Digg Deeper questions have been answered for this story yet.
Thrilled for my friend @dan_biderman to launch @EngramLab
Dan is a visionary who has consistently let seen around corners in AI
http://x.com/i/article/2069463677733142528
I've spent my PhD figuring out how a model learns from data, digging through piles of Spanish math problems, fanfiction, and the depths of StackExchange. Engram opens a new dimension: how a model learns from you - your world, how it evolves, and the threads that tie it all together. Figuring out how to do this well is a wide-open research problem. Join us :)
http://x.com/i/article/2069463677733142528
on a technical level it feels like we're automating post-training
we want to take a bunch of data and output a post-trained model - then iterate this process many times. continual learning by induction
the ideal algorithm is an interpolation of OPSD and RL with a lot of synthetic data
finally sharing what i've been up to! left phd end of 2025 and co-founded Engram.
there are a few startups in SF right making very different bets on the right way to train AI models. this is ours:
people want models that learn over time, remember details, adapt and interact like a person would
everyone gets a model. your model updates ~every minute. this is the world we're building. :)
congrats @jxmnop ! so many aie speakers announcing companies ahead of the conference
http://x.com/i/article/2069463677733142528
I'd followed the @EngramLab team's work for years before they joined together to start the company. It was clear they were each obsessed with a problem most of the industry hadn't fully appreciated yet: memory.
Engram gives an organization a memory its AI can reuse, so models stop relearning the same context on every query. The results hold up against frontier systems at a fraction of the cost.
They are out of stealth today with $98M raised. So excited to lead their Series A & partner with @dan_biderman, @EyubogluSabri, @realJessyLin, @jxmnop, @scott_linderman, and Chris!
https://www.kleinerperkins.com/perspectives/engram-giving-enterprise-ai-a-memory/
http://x.com/i/article/2069463677733142528
I'm excited to share what we're building at Engram! This team is incredible, and we're working on one of the most interesting problems in AI right now: how to build models that are tailored to each person and continually learn from experience. Come join us!
http://x.com/i/article/2069463677733142528
This looks like a cool startup working on continual fine-tuning of LLMs which are personalized to your data/context. Distilling tokens to weights makes a lot of sense for many reasons.
I'm excited to share what we're building at Engram! This team is incredible, and we're working on one of the most interesting problems in AI right now: how to build models that are tailored to each person and continually learn from experience. Come join us!
Memory is a huge challenge for frontier AI systems, and true personalization remains out of reach until progress is made on it.
Excited to see close friends and former lab mates go after these hard problems with new ideas @EyubogluSabri @MayeeChen @dan_biderman @scott_linderman
http://x.com/i/article/2069463677733142528
if you follow me, you know that
again, the AI labs are obsessed with building reasoning-native language models when they need to be building *memory-native* language models
- this is possible (the techniques exist) - no one has done it yet (no popular LLM has a built in memory module) - door = wide open
Today AI models can't delight users because they don't remember user behaviors and preferences.
Over the past 5 years I’ve gotten to know @dan_biderman, @realJessyLin, @jxmnop, and @EyubogluSabri, a research dream team. If anyone can solve the memory problem it's them.
http://x.com/i/article/2069463677733142528
also in general i've been really impressed with the amount of DEPTH there is to building models in industry, and working on larger projects over longer timespans
honestly the first model we trained was not very smart. a few clever ideas and so many little fixes/improvements later, the model is surprisingly intelligent, and i talk to it all the time
it feels good
i post about this a lot
this is insanely cool. engram is pretty much working on a model that can improve itself through every interaction across *all* users
which will likely be the single biggest capabilities jump since o1-style chain of thought reasoning
http://x.com/i/article/2069463677733142528
I'm really excited to be working with Engram to bring their models into Custom Agents in Notion. There's something magical that happens when the agent stops needing to search your workspace, and just knows it.
http://x.com/i/article/2069463677733142528
Engram is making a personal AI that really learns and internalizes everything you care about. It’s a different kind of lab with an amazing team put together by @dan_biderman @EyubogluSabri @realJessyLin and @jxmnop. Very excited for what they are building!
http://x.com/i/article/2069463677733142528
@MayeeChen it's been fun digging into data together! and seeing fun examples of generalization from weights
except that one time our model hallucinated that Sabri has a dog named Pearl. good thing we fixed that
I've spent my PhD figuring out how a model learns from data, digging through piles of Spanish math problems, fanfiction, and the depths of StackExchange. Engram opens a new dimension: how a model learns from you - your world, how it evolves, and the threads that tie it all together. Figuring out how to do this well is a wide-open research problem. Join us :)
@TheStalwart Joe you might not want to hear this but as we speak I'm scaling compute on your context. the model is studying Moby Dick, it's developing new theories of orality, it's watching Fed livestreams.
I wrote more about what makes Engram special and why this problem matters in my latest blogpost: https://www.amplifypartners.com/blog-posts/announcing-our-investment-in-engram-the-memory-dream-team
Today AI models can't delight users because they don't remember user behaviors and preferences.
Over the past 5 years I’ve gotten to know @dan_biderman, @realJessyLin, @jxmnop, and @EyubogluSabri, a research dream team. If anyone can solve the memory problem it's them.
@swyx crazy pic but yes come to AIE and I WILL GIVE MORE CONTEXT ON SCALING COMPUTE ON CONTEXT
congrats @jxmnop ! so many aie speakers announcing companies ahead of the conference
Incredibly bullish about the approach that @dan_biderman and @EngramLab are taking to make AI systems more intelligent. This will be a team to follow.
http://x.com/i/article/2069463677733142528
Been a fan of the work from many people on this team. Congrats on the launch! Exciting vision!
finally sharing what i've been up to! left phd end of 2025 and co-founded Engram.
there are a few startups in SF right making very different bets on the right way to train AI models. this is ours:
people want models that learn over time, remember details, adapt and interact like a person would
everyone gets a model. your model updates ~every minute. this is the world we're building. :)
.png){kind=avatar}
