Yutori
Reimagining how people interact with the web.
318 followers
Reimagining how people interact with the web.
318 followers
At Yutori, we're building proactive, always-on AI agents agents that can reliably execute tasks on the web, so you can focus on what matters to you.
This is the 3rd launch from Yutori. View more
Yutori Delegate
Launching today
Delegate is an eager AI agent that understands your context, never sleeps, never forgets anything, and loves busywork — research, coordination, monitoring, communication, admin, etc. Always-on, always yours. Throw tasks at it and move on. Come back to threads closed, research compiled, replies sent, follow-ups tracked.
Payment Required
Launch Team
Yutori
Hey all, @abhshkdz here from Yutori! 👋
We're excited to share Delegate — an agent you delegate work to and move on with your life.
yutori.com/delegate
With AI tools today, you’re getting a lot more done, but you're still stuck doing the heavy lifting — keeping track of every open thread, supplying the right context for each one, constantly context-switching, finding the perfect prompt.
Delegate is different. You connect your apps (Gmail, Calendar, Slack, Notion, Granola, etc.) and hand it tasks or a messy braindump, and move on.
Your Delegate takes it from there — orchestrating across apps, navigating websites, drafting replies, messaging people, researching the web, filling forms, creating slide decks and live dashboards — and more.
It's proactive. It schedules itself forward, checking back on threads, waiting on replies, picking things back up. Without you having to think about it again.
It's great with ambiguity too. You can hand it things where you don't quite know the plan yet, and it'll figure stuff out as it goes. If it hits a wall, it comes back with what it tried, what it thinks the issue is, and what it needs from you.
Building and using this, I've had to re-learn how I work with agents. Delegation asks for a different kind of giving up control. Not the back-and-forth you're used to with LLMs, where you're steering every turn. You hand something off and let it run.
It takes some getting used to. But once it clicks, it's hard to go back. I've caught myself reaching for it for things I'd normally just grind through — chasing down a vendor, digging up context before a meeting, coordinating a handful of small threads that each need a nudge.
We've been heads down on this for the past several weeks. A ton of love and sweat went in — really proud of what the team's put together.
Give it a shot and let us know what you think! Reply here with how you're using it, what's working, what's not.
So you are using computer use here - How are you taking care of the latency - does small language models help?