Simular

In another universe, you missed your kid's recital. Your mom's birthday dinner. That anniversary celebration with your person. In this one, you have 𝐒𝐚𝐢. The AI co-worker that does your computer work so you don't have to choose.

At @CalHacks 2026, we are thrilled to celebrate two exceptional projects that showcase the power of computer use agents (CUAs) and their creative, real-world applications of Simular's tech. 🥇 STING - A platform that detects scams early on, explains them in plain language, and provides evidence—going beyond simple flagging. STING showcases a novel use of Sai by simulating diverse user personas (elderly, Gen-Z, etc.) and leveraging @sai_borg's perception and computer control capabilities for product QA like website navigation, and optimization. The (one-person) team also used Sai to create their demo video. Congrats @vivianbuilds! 🥇 Tendly – An app that listens and reacts to elderly patients' needs, addressing a critical societal challenge: seniors often struggle to clearly communicate their needs while caregiver was resources remain scarce. Tendly generates TypeScript Simulang scripts and executes them directly on the desktop via Simular — no caregiver required. Congrats Caden Lewis, Henry Bonomolo, Seth Rojas, and Deven Mittal! Stay tuned for the demos. Thanks to everyone else who participated. Until next time!

getting saved by @sai_borg to get this @CalHacks ai hackathon done, any other use cases? Messages and usage is great tho

Built something huge in the past 24 hours at Cal AI Hacks 2026 🐻 STING — Scam Tracking & Intelligence Network Guard — helps families catch scam pages, suspicious links, fake bank alerts, voice scams, and handoff the evidence before someone gets hurt. Video made with @sai_borg

Judging time at #Calhacks - who's gonna win our $500 prize? 👀

Had a blast at day 1 of @CalHacks - truly inspired by all the ideas and questions from the hackers! Pictured: Our co-founder and CTO Jiachen Yang speaking at the opening ceremony and the Sai workshop; Sai's limited edition swag; our team @ZeningChen42844 and Minh demo'ing and brainstorming with hackers on how to build with Sai, Simulang, and Agent S. We'll be announcing the winner of the Sai competition tomorrow ($500 prize!). Your last chance to join: sai.work/prize Stay tuned 😎

Don't you wish you could rescind your likes on dating apps? Well... most don't let you. But LinkedIn does. It's a growth hack - except doing it one by one is painful. So @@ZeningChen42844 delegates Sai to withdraw invitations older than 3 months, and resend them with better messages. 🤓 Next on his wish list: get Sai to help send better likes on dating apps.

just gave my keynote at EDB's AI for Corporate Functions event talked about agents that don't just chat they actually use the screen, mouse, and keyboard. coworkers, not chatbots. the panel after (singdata, sierra, FWD, singapore growth capital, moderated by EDB global ventures) ended on the question every enterprise eventually asks: what actually matters for adoption? the answer wasn't vision. wasn't benchmarks. wasn't the model. it was ROI startup founder, hyperscaler, investor, group CDO all converged on the same word. that's the bar. and honestly, it should be. flying back energized. more soon from @SimularAI

"AI can't handle my LinkedIn networking.' That's what a sales professional told us during NYC Tech Week. "I'd love AI to arrange my connection requests and networking opportunities. But I doubt AI will be able to do that." Turns out @sai_borg can. LinkedIn hates bots, but Sai isn't one. Sai is a computer-use agent that works exactly like a human would (reads profiles, personalizes messages, finds the right people). Nothing happens without user approval and it knows when to ask for human control. AI doesn't replace the relationship building. It just handles the grunt work of getting us there.

We will be back again soon!

Thankful to be hosted by @SGglobalnetwork and @SGInnovate , speaking at the @XooglerCo meetup today including a fireside chat with Erica Ding and a sharp, curious crowd of ex-Googlers turned founders, operators, and investors. We went deep on where computer use agents are really headed and what it takes to make them work inside real organizations. Three ideas that sparked the most conversation: 🔹 The token economy of agents isn't at its endgame yet. Right now agents are token-hungry and expensive. Over time, that has to invert, well-built agents should require less token usage to do the same work. Efficiency is where this is heading, and it changes the economics entirely. 🔹 Repeatability is the key to business adoption. A demo that works once is a party trick. What enterprises need is an agent that does the same task the same way, every single time. Reliable, repeatable execution, not one-off brilliance, is what unlocks real deployment. 🔹 Early adopters get a compounding advantage. Whoever adopts agents ahead of the curve doesn't just move a bit faster, they pull away exponentially. As agents handle more of the work, the gap between early movers and everyone else widens fast. In a competitive market, timing is leverage. Huge thanks to the @SGInnovate @SGglobalnetwork @XooglerCo community and @ChristopherFong for having me, and to everyone who kept the conversation going afterward. The energy in that room was a reminder of why I love building in this space. 🚀 Building toward agents that use computers, repeatably, efficiently, and at a cost that scales, with @sai_borg @ @SimularAI . #Xoogler #AIAgents #ResponsibleAI #Startups #Simular #AI

How many hours a day do you spend moving your cursor? @lihautan walked the room through what changes when an AI agent uses your computer the way you do: clicking, typing, navigating real software - not just calling APIs. Great crowd, great questions. Thanks for having us, #AIContext 🙏

Honored to speak at @WorldBankGroup today about 100x organizations with agents that use computers. What struck me most were the questions from the room: 1. Cost is the first barrier. Many existing agents are simply too expensive to deploy at the scale organizations need. Bringing that cost down is what turns AI agents from a pilot curiosity into something every team can actually use day to day. 2. Safety is the deciding factor, and it points to on-premise. For governments and enterprises, the question "is it safe and under our control?" On-premise deployment keeps sensitive data and operations inside the organization's own walls, which is what makes adoption possible. 3. Transparency is what unlocks the everyday worker. Because GUI agents work on the screen - clicking and typing the same apps people use - staff can see exactly what the agent is doing. That visibility builds the trust that lets normal workers, not just engineers, put agents to work. Grateful to the @WorldBankGroup team and fellow delegates for such an open, practical exchange. Building toward an AI-empowered organizations with @sai_borg . #AIAgents #ResponsibleAI #Singapore #WorldBank #Simular

Hackers of @CalHacks, get a chance to win $500/teammate! We are excited to be part of Cal Hacks at UC Berkeley June 20–21. Build with our computer-use agent (CUA) @sai_borg or our scripting language Simulang - every hacker gets UNLIMITED credits for 2 days. 🏆 Rules for participation: sai.work/prize 🎯 Join our workshop "Vibe Hacking" for tips & hacks 📍 Find us at the booth. Founded in 2023 by former Google DeepMind researchers, Simular is backed by investors including Felicis, NVentures (Nvidia), South Park Commons, Basic Set, Lenny Rachitsky and more. Go Bears! #CalHacks #Berkeley #AI

You are wasting your tokens — and you don't have to. Your agent pays full price every time. No learning. Same task, same amount of token spent. @sai_borg learns. Run it once → cheaper every run after. More reliable + more efficient + more affordable. 🌱 Watch our CEO @angli_ai explains ⬇️

If you missed @angli_ai on panel — here's what mattered: Many agents went viral. But no enterprise can actually deploy them. No guardrails. No audit logs. No security review path. That's the opening. 👀

My @sai_borg edited the slides. I didn't open my laptop. Here's what makes this different: 👉 AI slide tools have always had one flaw — you can't really edit the output. You're stuck prompting back and forth forever. ✅ Sai generates the deck and uploads it directly to your Drive. Every element is yours to edit.

It's 10am on Friday and @ZeningChen42844 has checked out already... watching @sai_borg work for him. Plus, Sai is getting a parallel view of agents in action and Pixel art workspace 👾

Two days at @superai_conf, 10,000+ in the room, one Main Stage moment we won't forget Thank you Singapore🇸🇬

Agent Hour 🇸🇬 — done. A room of founders, engineers, and operators packed into @WeWork for the realest fireside of Singapore AI Week: Agents, Honestly — what's working, what's hyped, what's next. @angli_ai · @lionelsimai · Jay Teo — moderated by Freddy Lim. The gap between demo works and customer pays got real. So did the room. Huge thanks to our co-hosts Claude SG · The AI Capitol · First Prompt · @buildclub_ · @AgoraIO — and to @sos_266, @AnnieLiao_2000 & team for pulling it off.