Authenticating browser agents into websites is a huge pain on mobile devices
Either you send your password in plain text to the agent, or type directly into a remote browser live streamed to your phone
In this week's Buoy update, @riccqi made it possible to sign into sites using a native webview, complete with password and OTP autofill just like you’d expect
Your login session is then synced to our cloud browser so the agent can keep running even when you close the app!
We’ve completely redesigned Buoy for every release so far. In this week’s update we finally found the right form factor
V1 took inspiration from Safari’s tab UX. We used gesture based navigation to make creating new threads and swapping between them feel effortless.
V2 was akin to the UI of a to-do list. This design optimized for executing long-running async tasks in parallel.
V3 feels more like texting a person in iMessage — it’s a single continuous chat thread. It pairs naturally with long-term memory and encourages you to “teach” the AI when it lacks information or makes a mistake.
Buoy's browser agent now learns about you and improves the more you use it.
We designed a multi-level memory system to remember your preferences, store login sessions, and keep notes on the most efficient way to navigate websites.
And if Buoy thinks you will repeat a workflow in the future, it automatically creates a "skill" that can be reapplied in a single tap.
We’re working on creating the best agentic browsing experience on iOS.
Buoy can operate a browser just like you do, so it can do more than just find information. It can sign into websites, make purchases, book reservations, do research, and more!
Try it out today at https://t.co/VYgpsT8kmO
Buoy is the best way to automate browser tasks while you’re on the go.
* Designed to run in the background
* Lets you directly control websites when needed
* Auto-learns skills
* Persists website login across threads
Sign up to be an early tester:
We made a calendar app that gets smarter the more you use it.
It's more than a record of events. It notices your habits. Learns your scheduling preferences. Eventually, it will manage your calendar automatically, just the way you would.
Try our beta for an early preview!
literally just found out about https://t.co/2YAODiXcaV and I can already see how this will be absolutely game changing in terms of my work efficiency.
@ronithhh you gotta add an custom ai agent feature so we can select agents according to the task we are working on
Here are 3 reason why I feel like making the switch from chatgpt to iris @ronithhh @interfaceclubhq :
1. Accessibility: Iris appears at the corner of my screen with a simple key press, allowing me to view my existing work seamlessly.
We've been working on adding widgets to Iris — mini apps that appear within the chat thread.
Today we're launching the first one, the "proofread widget."
Voice mode on the ChatGPT mac app demos well, but the UX isn't great.
On desktop, ChatGPT shouldn't speak back to you. It's faster to read text vs listen to it
And auto-responding when it detects a pause in your voice is neat, but means you don’t get time to think while speaking
So we made our new voice feature in Iris simple. It's toggle on, toggle off. You get text output instead of a fancy AI voice. It's not as demo-able, but far more useful (but don't take my word for it, try it and judge for yourself! 🙃)
NOW HIRING AT @INTERFACECLUBHQ!!!
We're creating an AI layer on top of macOS that helps you get things done faster.
Looking for founding team members:
• A native SwiftUI/AppKit engineer
• An AI infrastructure engineer
(1/4)
Meet Iris — a macOS app that makes it easy to multitask with AI.
Over the last year, I’ve been exploring how we can evolve the desktop OS. This is the first small step :)
You can download Iris now at https://t.co/KALWfNCfQ6