Simranjeet Kaur

AI memory @mem0ai 🧠
Spent years building communities and relationships for brands. Now I work on the thing that lets AI do the same, remember you!

[Simranjeet Kaur @KaurSimranjeet9]

I spent 5 years preparing for India's civil services exams. Then 3 years managing online communities for an FnB startup, a fintech, and an ecom company. Now I talk about AI memory infrastructure.
Here's why that's less random than it sounds:

[Simranjeet Kaur @KaurSimranjeet9]

high time you breakup with all access to your agents now, with Gateway

[mem0 @mem0ai]

Everyone's racing to give agents more access
Nobody's asking what happens the week after

Try Gateway now: https://t.co/ISjg4Wcgc5

[mem0 @mem0ai]

Introducing Gateway.

Your agent needs to check a payment. You hand it a key that can also refund it.

It needs to draft a page. That key can archive every page.
Not anymore.

Gateway gives every agent one key with only the tools you grant. MCP servers and your own APIs, same key. Change the grant when the job changes. The agent never sees your real credentials.

Beta customers already run over 95,000 calls each through it.

Watch it work ↓

[Simranjeet Kaur @KaurSimranjeet9]

long awaited solution for the agent access issue is finally here -

[mem0 @mem0ai]

Introducing Gateway.

Your agent needs to check a payment. You hand it a key that can also refund it.

It needs to draft a page. That key can archive every page.
Not anymore.

Gateway gives every agent one key with only the tools you grant. MCP servers and your own APIs, same key. Change the grant when the job changes. The agent never sees your real credentials.

Beta customers already run over 95,000 calls each through it.

Watch it work ↓

[Simranjeet Kaur @KaurSimranjeet9]

@mem0ai Never thought AI memory can be explained this easily. 🙌

[Simranjeet Kaur @KaurSimranjeet9]

What's the one fact you'd want your agent to never forget?

[mem0 @mem0ai]

Your Hermes agent has a 2,200-character memory.

Its notes live inside the instructions it loads at session start, so the space never grows. Hit the cap and the next thing it learns gets rejected - to remember something new, it has to delete something it already knew.

Mem0 plugs in as a memory provider and runs alongside the built-in files. The notes stop competing for space.

Watch the meter hit 100% here -

[Simranjeet Kaur @KaurSimranjeet9]

check how your agent's memory works before you ship it, use Playground by mem0!

[Sebastian Buzdugan @sebuzdugan]

i told my ai agent i'm vegetarian and allergic to nuts.

then i killed it.

started a completely new session, and it still knew.

no chat history. no giant prompt.

here's how ai agents are getting long-term memory ↓

[Simranjeet Kaur @KaurSimranjeet9]

@ycombinator @physical_int @chelseabfinn After artificial intelligence, it is an era of physical intelligence it seems

[Simranjeet Kaur @KaurSimranjeet9]

@justinskycak How should one do this is the question? Where to start?

[Simranjeet Kaur @KaurSimranjeet9]

Follow along if you're building AI products and want them to actually remember your users. 🧠 Stay tuned!

[Simranjeet Kaur @KaurSimranjeet9]

I spent 5 years preparing for India's civil services exams. Then 3 years managing online communities for an FnB startup, a fintech, and an ecom company. Now I talk about AI memory infrastructure.
Here's why that's less random than it sounds:

[Simranjeet Kaur @KaurSimranjeet9]

I'm someone who's spent time thinking about how you get systems, human or AI, to actually remember and relate to people. That's what I'll write about here: explained for builders.

[Simranjeet Kaur @KaurSimranjeet9]

mem0 on @Cursor ftw 🙌

[mem0 @mem0ai]

Mem0 plugin is now listed in Cursor Marketplace: add it once, and every chat can recall decisions, fixes, and conventions from across your whole codebase history. Not just this session.

The plug-in is here: https://t.co/GBWeGGKKeg

[mem0 @mem0ai]

We tested whether AI agent memory has to get messier over time, or if it can clean itself up. Here's what happened when we ran Mem0's new Dream feature on a Hyrox athlete's training log.

[Simranjeet Kaur @KaurSimranjeet9]

redundant memories taking more space? Dream feature ftw!

[mem0 @mem0ai]

Most active projects carry a few memories that quietly contradict each other. We built something for that!

Introducing Dream, the background layer that keeps mem0 memory clean, current, and true.
→ Merge: facts that repeat get folded into one clear memory
→ Supersede: the old fact stays as history, the new fact becomes the truth
→ Synthesis: not one memory — a pattern built from many, traced back to each one

Excited to share that we're live today on Pro and Enterprise!

[mem0 @mem0ai]

We cut vector search latency by 70x, from 8–14 seconds to ~110–120 ms. The biggest unlock? Moving vector search to @turbopuffer and keeping Postgres focused on what it does best.

If you're building memory infra for AI agents, this one's for you: https://t.co/bnpajkpneG

[Simranjeet Kaur @KaurSimranjeet9]

excited on what's coming up!

[mem0 @mem0ai]

We tested whether memory actually saves tokens in Claude Code or if that's just a talking point!!

13,700 tokens vs. 445 tokens

Here are two experiments we ran on the same repo, one with Mem0 and one without.