The Verifiable Data Platform for AI & onchain finance.
Walrus makes every byte provable, programmable, & always available.
Built by the team behind @SuiNetwork.
June was all about memory. 🦭 Here's what went down:
🎙️ @kostascrypto went on the ETN show arguing your AI memory could soon be worth more than your entire online footprint, emails, search, and social combined.
🤖 CARA, @AethirCloud's crypto AI agent for traders, KOLs, and researchers, is integrating Walrus Memory.
📊 Walrus Memory hit 38K memories registered in its first two weeks, ~2K agent owners, and around 7% daily growth.
👨💻 @cole_medin wired Walrus Memory into Claude Code, Codex, and Pi with one prompt, encrypted and owned by him the whole way.
🛠️ Walrus Sessions 5 went live: the Walrus Memory Prompt Jam, a $2,000 WAL pool for the best prompt that makes a Walrus Memory agent genuinely useful.
🎤 We joined the Sui Roundtable X Space to talk Walrus Memory, and teased the UI layer coming later this year that turns uploaded files and videos into agent memory.
🤝 Big congrats to our partner @AlliumLabs on their $40M raise, one of the few teams turning onchain data into something institutions can actually use.
📣 @Alkimiexchange's research with WPP reaches past adtech to show how agents without shared state drift apart, even rational ones.
🗓️ Sui Basecamp is coming to Singapore, October 7-8 at Marina Bay Sands with @token2049. We'll be there building portable memory for agents on Sui.
June was the month memory got real. July goes FURTHER. 🦭
Great breakdown from @KhalaResearch on how @WalrusProtocol is solving the AI agent memory bottleneck.
Excited to see @inflectivAI featured as part of the ecosystem building persistent, structured intelligence.
Happy 4th! 🇺🇸
The long weekend probably just means more time to ship.
But here’s some advice for young builders: go touch grass. 🌱
You can’t work at full speed every day. So when everything’s merged, shut the laptop and go be a person for an afternoon.
Your agents don’t get that luxury though. They keep forgetting. Walrus Memory gives you and your agents long-term memory, so you’re not starting over every morning.
Rest up. We’ve got the remembering covered. 🦭
Agents got smart fast. Remembering is the part that's still broken.
@KhalaResearch's latest report explores agentic memory debt: the growing cost of agents losing context, repeating work, and starting from scratch as deployments scale.
Read about where agent memory is headed👇
It’s now obvious that memory is one of the biggest bottlenecks in AI
But the next question is whether that memory can be trusted
“Context is the new bottleneck” - Jensen Huang at CES 2026
He’s right, but only half right
Every AI agent deployed today still has the memory of a goldfish because context windows are stateless by design, starting each session from zero
That is unfortunately how LLMs work
But storage capacity is only half the battle, the other half is trust
Current memory providers can write memories to docs, databases or internal logs, then ask enterprises to trust that record later
The problem is obvious: one altered memory corrupts every decision that relied on it
As agents move into regulated workflows, this becomes a real issue
The EU AI Act requires high-risk AI systems to maintain automatic event logging and traceability, meaning enterprises will need proof of what data an agent accessed, what it remembered and whether that record was tampered with
Our latest report covers AI memory and how Walrus tackles this problem with portable, verifiable and programmable memory for agents
The full report is in the next post below
June was all about memory. 🦭 Here's what went down:
🎙️ @kostascrypto went on the ETN show arguing your AI memory could soon be worth more than your entire online footprint, emails, search, and social combined.
🤖 CARA, @AethirCloud's crypto AI agent for traders, KOLs, and researchers, is integrating Walrus Memory.
📊 Walrus Memory hit 38K memories registered in its first two weeks, ~2K agent owners, and around 7% daily growth.
👨💻 @cole_medin wired Walrus Memory into Claude Code, Codex, and Pi with one prompt, encrypted and owned by him the whole way.
🛠️ Walrus Sessions 5 went live: the Walrus Memory Prompt Jam, a $2,000 WAL pool for the best prompt that makes a Walrus Memory agent genuinely useful.
🎤 We joined the Sui Roundtable X Space to talk Walrus Memory, and teased the UI layer coming later this year that turns uploaded files and videos into agent memory.
🤝 Big congrats to our partner @AlliumLabs on their $40M raise, one of the few teams turning onchain data into something institutions can actually use.
📣 @Alkimiexchange's research with WPP reaches past adtech to show how agents without shared state drift apart, even rational ones.
🗓️ Sui Basecamp is coming to Singapore, October 7-8 at Marina Bay Sands with @token2049. We'll be there building portable memory for agents on Sui.
June was the month memory got real. July goes FURTHER. 🦭
2,000+ teams build on IPFS through @LighthouseWeb3, and now they can upgrade to Walrus without changing how they build.
Keep the same CIDs, gateways, and SDK, but gain high-performance retrieval and verifiable data availability.
Nothing to migrate. Nothing to rewrite.
Learn more 👇
@Kalshi Worth noting alongside this: a lot of compute gets spent on agents re-deriving context they already had once.
Memory that persists is one of the few things that actually reduces the load instead of adding to it.
@CoinMarketCap Agentic payment systems are coming fast.
The piece that's still missing...those agents need memory that persists and can be verified, or they can't be trusted with anything that moves money.