Supermemory has raised $3M in pre-seed funding to build the best memory engine for agents. Agents needs to be personalized. They need to learn and grow with the user. They need memory. And that's the future we're building today. https://t.co/8fhhJ6tqHt

@supermemoryai docs appreciation post!

We're a team of two full time engineers (INCLUDING ME) and 2 part-time researchers right now. Growing too fast now. So I'm finally hiring! Looking for cracked, irrational thinkers and doers who love hard challenges. should be fun to be with Remote OK. In person (SF) preferred!

Supermemory is hiring for founding engineering and research roles. Apply here

A deep dive into MRL (Matryoshka Representation Learning) and how we use them in supermemory

https://t.co/7oSxpsPP6J

🔖 Bring your organized 𝕏 Bookmarks to supermemory and finally make those tweets useful that you've been saving for years Now available via supermemory chrome extension. Link below

Some customers faced degraded service from 1:17 PM to 1:45 PM PDT, with increased latency and request timeouts. We managed 2800 RPS, but unexpected traffic caused issues. The problems are now fixed. Full report here:

https://t.co/wZL9IFZIPL

You can use supermemory with no-code tools like @n8n_io, @zapier and basically all others. We wrote detailed step-by-step docs on the same. Below are some examples of how you can utilise supermemory to supercharge your workflows. See how it's done below 👇

it sucks when someone gets the details wrong While I can't fix the media, @supermemoryai can help AI remember the details about you. in real time. it's time to personalize our experiences with AI. it's supermemory time.

Supermemory listed alongside legendary companies like slack and cursor, front page on @planetscale!! We're also the newest company in the list 🥹

so awesome. Memory is useful in every single AI product out there. So when are YOU going to add personalization to your products?

Introducing: User personalization (profiles) in @supermemoryai It's the era of personalized AI. In just a few lines of code, you can now add truly magical, native memory to your AI app. How? Let me dive in, and why this is different from anything else that exists.

Imagine ANY model you use in @aisdk has your user's memory built in powered by @supermemoryai ❌ No more Toolcalls just to get user's memory COMING SOON 👀

Learn more about it here: https://t.co/yqxiqaarAr

That’s why we built Supermemory to handle both patterns in one system: - Documents (for RAG-style retrieval) - Memories (for contextual, user-specific understanding) A unified API that lets you query across both seamlessly So when you ask: “What phone should I recommend?”

Why RAG fails as memory: If a user says: Day 1: “I love Adidas sneakers.” Day 30: “Adidas broke after a month.” Day 31: “I’m switching to Puma.” Day 45: “What sneakers should I buy?” RAG will keep surfacing Adidas-related results because it’s only matching text. Memory will

Memory answers a different question: “What do I remember about you?” It’s stateful, temporal, and relational. Instead of just finding similar text, a memory system builds a graph: - Entities (users, products, concepts) - Relationships (preferences, ownership, causality) -

RAG (Retrieval-Augmented Generation) answers the question: “What do I know?” It works well for static information like documentation, specs, FAQs, and research. You embed a query, search a vector database, and retrieve the top-k results. It’s stateless. Each query stands on

Most developers confuse RAG with memory. They’re NOT the same thing. And using RAG as a substitute for memory is why agents keep forgetting important context. Let’s break this down: