Every day at Auki. 🫡

Part 2 of this demo: After creating the action generated by Cactus Insights, store employees will get a notification to do the task with navigation and instructions. See how Cactus starts to become the brain for store manager robots? 🧠

Thesis: robots will be infrastructure. Step 1: Deploy robots (<2 min on the real world web) Step 2: Allow others to add capabilities to our robots, so that multiple AIs can co-embody the same robot Listen to Nils talk through our strategy in last week's community update. 👇

Today retail robots can guide customers to products, as Terri MechKenna demonstrates below. That and other perception tasks are valuable enough that we can deploy robots already. Later someone else may add manipulation capabilities. That's the idea behind robots as infrastructure.

Recently got to know #dubai a lil bit more and also to understand a bit more what is Dubai actually made of. What i found out is that they bet huge on technology and the future. Not only do they bet on it, but they also envision and implement among the pioneers. Keeping this in

"Smart money" is getting smarter when it comes to investing in robotics. 🧠 World models and data collection were well funded in 2025, but this year money is flowing to robotics companies that can actually deploy. And we can already deploy robots that do perception tasks.

- Co-exhibiting with Bruegmann (huge shelving manufacturer) at Euroshop, the world's largest retail conference held every 3 years; Day 1 was exceptional - Discovered a 9 fig problem we can solve - Took American client to visit our Swedish client - 6 Booster K1s arrived! 🤖

Retail is one of MANY industries the real world web can disrupt. We focus on it because we're building Cactus for retail, but the same principles apply elsewhere. Perception robots can also do valuable work in agriculture. A new partner is tackling this for vineyards. 🍇

Planogram management is coming soon to Cactus. 🌵

Retail robotics is entering a pragmatic phase. Guiding customers to products already creates measurable value. Perception alone justifies deployment. That changes the timeline. $AUKI Once robots are in stores performing high-value perception tasks, the foundation is set.

Being best-in-class at mapping stores for robots doesn't mean we stop finding ways to improve. Instead of scanning every product, eventually retailers will just import planograms, edit them, and push live. Hours of scanning reduced to minutes. Here's an early demo.

CactusXR powered by $AUKI Auki is turning physical retail into programmable infrastructure. Retail mapping is shifting from hours of robotic product scanning to instant planogram imports that can be edited and deployed live in minutes. Instead of slow, manual environment

Physical stores are finally getting smart data that drive results Auki lets retailers use their phones to map exactly where products sit in the store and connect that map to sales data. Now, it’s not just about which items sell well. You can see exactly where on the shelf

▪️ $AUKI ▪️ $DMTR ▪️ $DEXTF ▪️ This market is a test for the ages I’m slowly scaling on surging sectors Lower market caps that are shipping Those showing rising social dominance When the next bull market emerges… They are the ones that will move fast:

Auki Community Update Feb 28 x.com/i/broadcasts/1…

Doesn't feel good to say "we told you so" but we told you so. If you're wearing Meta's AI glasses, someone is watching. “In some videos you can see someone going to the toilet, or getting undressed. I don’t think they know, because if they knew they wouldn’t be recording.”

- Nils is finally back in HK, will host Papers & Beers night to bring more of the robotics community to our lab - Cactus demos: planogram builder and floor-based heat maps - New robotics partner: Bracket Bot - OEMs offering us free robots - Reconstruction & CV updates

When robots take over stores and homes, they’ll need a shared 3D map of reality. AUKI is building it - decentralized.

Nils finally admitting to the real reason he started a robotics company.

Even a thousand years from now, AI will still need hyperlocal data about the world. Sensor DePINs are never going away. Catch up on today's DePIN Coffee Hour podcast! Nils shows up at around the 25 min mark.