Launch HN: Muddy (YC S19) – Multiplayer browser for getting work done

252 points by lele0108 ↗ HN
Hey HN! This is Jimmy, Ron and Austa from Muddy (https://feelmuddy.com/). Muddy is a browser for work that automatically keeps project files organized in the same place where you use and share them. Here’s a demo: https://www.youtube.com/watch?v=tZr49aN3sjQ. Download and try it out here: https://feelmuddy.com/.

Building together in the past, we were incredibly frustrated with how much friction there is to get anything done on our computers. I was losing time everyday digging through chat logs looking for that one important link or breaking others out of flow by asking where something is.

Web apps promised to help us get more done—and they do, but each in its own silo, so there’s still a ton of redundancy to deal with. Every app has its own way of organizing files, its own notification inbox, its own search system. Conversations live everywhere and there isn’t a single view to see everything about a project. Remember when files simply lived in folders rather than the “cloud”?

We started dedicating time to organizing our files in shared docs and limiting new apps we used. This helped – but the second we didn’t stay on top of organization, links became stale and things got messy again.

Muddy started as a hack week project we built for ourselves—a single place to use web apps with others, but personalized for each user automatically. Everyone gets their own view for every project, designed around how they work.

Muddy users work on projects in spaces, which are like automatic tab groups. Users share apps (any site works—a Github PR, Figma file, Trello board—whatever you want) into the project’s shared timeline and Muddy automatically opens relevant tabs for you. It’s a single click to open up all the apps you need for the project.

Under the hood, Muddy works in the background to keep track of the timeline and uses a LLM to continuously organize apps and keep everything on to date. It considers signals like the popularity of a file, naming conventions, and conversations to figure out what’s relevant. So everyone is presented with an updated list of important tabs, without anyone lifting a finger. Our actual browser is based on Chromium.

When you need to revisit something from weeks ago, you can rewind the project timeline to that point in a single click. Apps open up in the timeline so you’ll see your files right away. For sites that don’t have built in collaboration features (like documentation), Muddy lets you do annotations directly on the website.

Projects sometimes get big and need to be broken up. Across all your spaces, Muddy can answer questions like ChatGPT, cite your files as sources, and return apps directly. This is possible since Muddy’s AI shares your browser and can use your authenticated apps locally (with privacy in mind).

Other browsers like Chrome and Arc focus on solo productivity with sharing as a bolt-on. We think productivity depends on how well you can work with others, and should be the first class consideration. And doing organizational work manually is unsustainable.

Muddy will have paid subscriptions for teams with additional features like shared passwords, team organization, custom shortcuts, and SSO management. Those aren’t built out yet and the base product will be free. No part of our revenue will come from data monetization.

We’d love for you to give Muddy a spin! You can download Muddy for Mac or Windows on our website and add others once inside: https://feelmuddy.com/. We’ll be around to answer questions and look forward to any and all feedback!

113 comments

[ 2.4 ms ] story [ 211 ms ] thread
Congrats on the launch. I recently was trying out Microsoft Edge's built-in Workspaces feature which allows multiplayer collaboration. What would you say are the key differences?

https://www.youtube.com/watch?v=w5AX_HvtYfI

The main difference is that Muddy automatically creates and updates those shared tabs for you based on what's being added to the project timeline (multiplayer feed of apps/websites).

We used some similar tools like Workona in the past but without constant maintenance, links would get stale and we'd abandon it. Wanted something that did that for us automatically.

(Also we support: website annotations, team presence, and letting you rewind a project's timeline back to any point in time)

(comment deleted)
So, despite being based on Chromium, there's no Linux builds? Would like to try this but me and my entire team uses Linux.
Hear you! Coming soon. There are some OS specific patches we made to Chromium that make this less straightforward
That is really unfortunate, because it removes the only advantage of using the web to begin with. Hopefully you guys are able to find a solution.
We have Ubuntu and Fedora users through Wine. Not ideal I know, but enough to try out unless you have a super custom Arch setup
I don't see what having an arch setup would have to do with running it through wine, but running your main browser under a heavy compatibility layer is practically unusable in my book. I need my browser to be able to be invoked in no more than a couple hundred milliseconds, while wine programs can sometimes take 5 to 10 seconds to launch.
Yes, the lag is the main problem. Certain distros and setups don’t work with wine. Mine doesn’t, but I’m sure it’s a tractable problem.
This is really cool! Congrats on the launch!

I think my usage of figma,sheets,etc. is 90% single player, until the moment of sharing my (maybe unfinished) work, where I go through an intense period of collaboration with others for an initial review, then tails off, and becomes async.

I can't see myself using muddy for the single player part, but it sounds interesting for after that initial intense collab process. Especially if the process includes multiple apps, as opposed to a single design review in figma etc. I find the longer running async collab is when I get the most scatterbrained across apps.

I felt this pain a lot. Never really resonated with all-in-one apps so we always had a few places to check for every project. Figuring out the right links was hard and prone to distraction.

In a way, we like trying new software. Downside is each app has the need to build their own slightly different file system, which makes finding things extra challenging. Wanted to solve with Muddy.

I can see why managers would love this, social pressure around doing work and working faster.
I can't say we haven't heard that. Surprisingly, it's usually more of a TL or most Senior Eng who seems to like those presence features. A lot less cat herding across apps for them.

We have less traction with PMs. Most of the feedback from them implies they like being bombarded with Slack notifications even if they say they don't.

I'm not sure this would be useful to me, because I don't switch between various projects that would benefit from different tab groups, but sounds like it could be useful! Good luck.
Is there a Linux version?
Not yet, in the pipeline. Depending on your distro (Ubuntu and Fedora) have users through Wine.
What's the story around getting into YC in 2019 and then launching in '24? I'm guessing there were some interesting pivots along the way ...
You are right. We came in letting hosts open experience stores in Airbnbs (try on an Oculus at a hosts home). Difficult to attribute our sales so we moved on.

COVID hit and we had the itch to look at the browser in a different light. Building a browser seemed intimidating...but it was quarantine. Long story short, built a few different ideas and here we are :)

Building a browser does seem intimidating! What technical considerations made you finally take it seriously? Any advice for folks trying to do the same or taking on another intimidating project?
Started just by building Chromium and Brave and poking around in there and making small changes.

Chromium is huge but has terrific documentation (once you find the current copy) and Google hosted code search. Digging through crbug.com often helped point us in the right direction.

2 unlocks that made us confident to pursue this project:

1) Repeatable way of patching Chromium and keeping up with upstream changes 2) Writing UI in web technology instead of C++ Views toolkit.

This is incredible, 5 years of not giving up, kudos. How big is your team and how did you make your money go this long?
Team is a handful, fluctuated over the years but never more than 8.

We’ve always treated everything we do as an experiment. We’re not trying to keep anything alive for the sake of it.

You can actually check out all the material from the pivots inside the app. There's a walkthrough in the Getting Started space when you first set up.
I spent a little time googling around but even wayback machine is only showing results in 2024 for the domain.

Was there a different project that was being worked on previously? Or is this the only product being launched since YC S19?

Previous projects of ours include: sail.online nototo.app

This is the first one with real traction and retention, so it's the first time we started to share on HN. Wanted to make sure it was GA and not a waitlist before we came here

The feature for sending messaging and posting comments to a tab is some pretty clever and creative UX. Seriously next level future stuff and congrats for just coming up with the concept.

I like that it's all timeline based. For my use case, we currently use Front email thread and then link to a Shared Dropbox where we post everything (including links to like a Google Doc or webpage). I think having chronological bookmarks like you do would be clearly better. I also know many people who use Google Groups and Google Doc to document progress too -- which I think would be insane / nightmare but teams do it. You all definitely would solve that automatically.

Couple other notes:

- Whenever I screenshare with a team or others I see 1000 bookmarks or tabs on their browser. I could not imagine the nightmare of how that would impact my workflow or the timeline. Trusting AI to clean stuff up or hunt is not for me.

- I can tell you all have been heads down blitzing (dog in video, phone ringing in background of another) but I think a separate "Solutions" page where you tackle specific examples would be nice to see or browse.

- Maybe too much or not really your goal, but right now need some sort of client integration for an outside person. I can't imagine giving access to a client on a whim and training them on this. Instead, maybe automatic email integration where their emails show up in the timeline and can respond directly from there. Would produce a really great timeline for where things left off and when things are being communicated. Being able to sub-comment and share files/updates/things on Front on email threads is one of the most killer features for productivity and a team. Mixing this with what you all have could be even more next level. Again though, might not be the goal.

Congrats and best of luck! Big fan of people trying to tackle PM stuff and think you all are doing a great job.

Great idea for creating some specific videos for engineers, designers, etc.

Browsers are interesting since they can do almost anything but "you can do anything you want!" is intimidating for many new users.

We've given e-mail thought and it's certainly a door we are considering as we keep on building. Has anyone built a browser without thinking about an email client? :P

Integrating an AI tab organizer of which they’re very few and the chrome one is essentially garbage, would make this a definite purchase for me. The AI should organize my own tabs and the ones I share
yup, that's Muddy. One of our users described it as a self healing slack channel with tab groups.
Question: What were the technical requirements that necessitated a custom browser? That would really sink adoption for my team. Any plans to have a lite version that can work inside your favorite browser (even if plugin is required)?

Idea: your privacy section only talks about cookies and ads, but all my privacy questions were around the AI features that would use all our team's messages and work across apps as context. Would definitely cover that piece.

Most of the stuff we do that people seem to really like can't be technically or physically done inside of another browser's tab unfortunately. Chrome extensions are too limiting as well. We started off as a chrome extension inside of our previous company, and hit a wall pretty quickly.

For the LLM calls, the one's you currently see are patching calls to a variety of model providers. As long as they hold their end of the TOS you should be fine. Everything else is happening locally.

We will launch a self-host option for Muddy. Cool features like hosting the servers and build process yourself, custom icon and skin on the app, and other internal rules you can set up. Let me know if that's of interest.

Still not enough. You are using Chromium but your software might not be secure enough. You are dealing with sharing data all over the internet and part of your code is not opensource.

If we cannot inspect or reproduce what you are doing , we cannot trust our work with your browser. We might as well consider you are stealing our data.

If you launch self-host option , you should opensource it.

What we have for opensource alternatives are :

https://github.com/BrowserBox/BrowserBox https://github.com/m1k1o/neko

(comment deleted)
We are interested in your mixture-of-app-tabs for SaaSes with SSO, contained* in a secure MOAT™.

Offer Muddy to most people here, who DGAF (datafeed Google and Facebook).

Upsell MOAT™ and your self-hosted version (leveraging each of M365, Google, and AWS identities** for the browser profile, and hosted inside company VPC that also leverages that native IdP) to companies that would like to let employees use SaaS but haven't figured out how to put a circle around that.

If you are interested to discuss, reach out.

* See the inTune ecosystem where, building on MSFT notion of data container for their O365 suite, apps from arbitrary vendors can share a data-loss protected data space (see Zoom Workspaces for inTune). The way you are writing the browser yourself, you could, in theory, have an option that ensures a web app is using SSO or it can't receive data from the user, other than a white list such as the SSO IdP flow, search engines, and pre-SSO login pages of company's preferred apps.

** Start with M365 since 85% of businesses in U.S. market use it, the IdP is already in their Office user seat plan, supporting "Sign in with Microsoft" as well as "SSO" flows, which picks up most web SaaS companies need.

So does "local data stay local" or what? I mean, obviously Trello cards are stored in Atlassian's cloud somewhere and Slack convos are somewhere in Salesforce's series of tubes. I guess what I mean is, if "Muddy works in the background to keep track of the timeline and uses a LLM to continuously organize apps and keep everything on to date", does this mean Muddy ships off local data to some remote API to get it's work done? Or if I'm only ever working within my local network, is the data I'm working with inside the Muddy context as safe as my local network can make it?
An integration with current chat system (Slack, Teams, etc.) would be neat. Changing where the work happens is a big ask in my opinion.
I can't really imagine how integration with them would work.

I don't want to speak for the Muddy people, but I wager they probably think chat app is kind of miserable experience. This instead tries to have a superior offering of something cleaner, easier to visualize where things are, and then move that "slack/teams" type convo to specific comments and tasks.

Yeah, unfortunately there's no real way to deliver on a consumer experience on top of another browser. Craziest implementation I can think of is writing a MacOS app that's a wrapper, but obvious distribution/compatibility issues there.
Congrats on the launch - seems like a neat product. A few questions: 1) It seems like you iterated on a bunch of ideas that landed in this - what were some interesting features that the team was really interested in that you ended up killing? 2) What were the most non-intuitive hard technical challenges? 3) Have you successfully ended up having someone give up on Slack to use this?
1. Spatial canvas based browser (we called it Sail). Had people who loved it but their team refused to learn it. Spatial tools are just incredibly niche and difficult to get started (besides Google Maps).

2. Auto updating across Mac and Windows (in 2024!!). Google Omaha is hard to setup, Sparkle is jank on Windows. Throw in binary delta support and it's an oof.

3. Our team builds Muddy on Muddy and ditched Slack when product got stable. Few other beta users and their companies as well. Slack is super sticky and has some terrific workflows, but more "quality" conversations today happen natively in apps and almost all apps have commenting functionality. Just easier to talk next to the context. So a lot of Slack convo's become "where is X" and we think Muddy will stop the need for those questions all-together.

Yeah auto update on Windows is a mess. One way to get it is to use MSIX, in which case Windows will delta update your app for you, even when it's not running (Chrome style) but without the need for special servers. Unfortunately old Windows versions have a lot of bugs. My company sells a product that makes all that easy to deal with and works around the bugs but it was a hard slog to get it all working well.

Sparkle on Mac is great though.

Congrats on the launch!! Collaborative browsing is something I've been looking for a few use cases of mine. Excited to try it out.
Is there a listing of what data is sent back to your servers and how that data is stored/handled?

I work in a secure environment. I like the idea of this app, but leakage is a huge factor for my teams.

That aside, I've already downloaded a copy and plan to try it out in a non-secure environment. The concepts here look like a great idea. GL and thanks!

Nothing sophisticated right now unfortunately. But we're planning on releasing a version where all servers, models, and build CI for the app are self-managed. Let me know if that's of interest.
Looks very interesting. The chat feature almost looks like a good slack alternative.

I can imagine this being useful to organization where your chats and your work are in the same window.

That's the general idea. We found that if you break down the average team's project slack channels, it's a bunch of urls and threads about them. Just wanted to make that experience simple!
You can use Slack & Discord from browser tabs. That's how I roll

One Chrome window for

- work (slack, jenkins, bitbucket)

- personal (discord, gmail, twitter, HN)

- open source (github, docs, projects)

Each one is a mental box and are kept separate

Congrats on the launch! Awesome to see this release as an early user who was able to check it out. The shared workspaces and shared browser windows with context in place has been incredible for collaborating with folks. We have our Figma design, Notion doc, and Gitlab MR all in the same space so we don't have to go searching for each one independently or have them cross-linked to each other.
Congrats on the launch! One of the positive byproducts of the open web platform is that products like this are possible. Much of what I share in other communication tools are links to things on the web, but sharing ends up forcing a context switch cost on both my end and the receiving end. Love that Muddy is exploring this problem space since I haven’t felt like other neue-browsers have gone far enough in making browsing itself a more collaborative and in-context experience.
This is super neat. I would love to use this in a team setting if I could convince the org to pay for it.

I saw in another comment that it’s a patched version of Chromium. Are you using CDP for underlying communication?

Totally evokes the good memories I have of "Google Wave" as a way for folks to collaborate on rich documents. Super cool.
Google Wave for the web has been one of the more popular ways to describe Muddy. I didn't see it at first, but I'll take it :)
Google Wave was the first thing that came to mind for me as well. RIP

That being said, this looks like a nice spiritual successor to it!

I came here to say this -- couldn't agree more. Very positive feelings and this nails it (and adds some stuff, too).
Yep! Glad to see I wasn't the only one who remembered Wave and what it had the potential for.
Love this and am tremendously impressed by this team's persistence. Congrats on the launch!
The unsubscribe link in your welcome email is broken. It isn't a link.
Ty for the catch, correcting now. Seems to be for the invite and update emails?