It would be great if you can share your journey, and why you think it failed? And how you would do it differently if you start today?

Also this is not the first time I'm seeing a universal search startups fail, some even YC backed. It sounds like an interesting idea, but why is it failing, I can't seem to figure out, maybe not enough market or adaptation?

Honestly, the story isn't that interesting. When I started I barely new JavaScript. Over the course of development, I learned Rust and TypeScript, and grappled with so many edge-cases and intricacies of NodeJS, SQLite, Tauri, Tantivy, and many other technologies. It was a great learning experience for me.

The original idea emerged from my workplace where everyone has thousands of documents on their machines but can't recall what's where, so they always ask each other for files, and then end up with several duplicates. It's a mess. I thought a natural language powered high-speed, low-footprint app would be a great USP, especially when you compare it against Windows Search.

Turned out, everyday office people (not nerdy hackers) have their own set ways of managing and discovering files on their machines. A guy I met keeps ONE 1000-page Word document instead of hundreds of smaller Word files because Ctrl+F in a Word document is more effective than Windows Search! Plus, a lot of files are now moving online to Google Drive or are locked behind Microsoft OneDrive, especially if it's a corporate workplace. Corporate work machines also made it quite impossible to install an app with that isn't code signed. So my target audience was kind of priced out for me!

I didn't start out to make an 'universal search app' but I fell for its allure somewhere along the way. I wanted to make a simple desktop search app that _just works_. Kind of like Everything [1] for Windows but full-text, cross-platform and less nerdy.

In the end, what I got was an app that fits my needs perfectly and I am really happy with it!

[1]: https://www.voidtools.com/

Yep, you were onto something. Everything is so useful, almost vital now. If it were cross-platform and had a text-search option it would be perfect. I think the only other limitation is the columns for search and sort are but a handful. Though index and search would presumably be slower, so it would need to be optional.

But the question that I wonder is, why even cross platform, online search tools, which search your google drive, one drive, dropbox and show results in one place, even tend to fail,

I have seen few startups even YC backed that didn't make it, I wonder why it's such a hard barrier to entry, even being a common problem everywhere

The only product I have seen do well is Glean [1] but they are B2B so not exactly consumer-grade software.

I think the everyday person works better with the separation of concerns that comes with using multiple services. For me personally, I vaguely recall that I came across this file/info on WhatsApp or on email or on Google Drive. Then I go use the search built into those services and it works well most of the time.

Buzee really helped me with local files (of which I had several thousand) but if you don’t have too many of those, Spotlight and Windows Search combined with some pragmatic file management seems to do the job fairly well.

So the pain point isn’t all that painful and the inertia of switching to a new omnichannel search interface is much higher.

This is why I kept imagining Buzee’s _real_ USP to come from a layer built on top of the file search engine. Imagine a personalised LLM trained on your files and organisation patterns only. Or a service that creates structured data out of the mess of your files. Or a butler service that organises data from every single service you use (goes beyond files to HN, Reddit, Spotify, Netflix, Strava etc.)

We’ll probably get something like that in a couple of years with a strong vendor lock-in. A privacy-first open source alternative seems difficult.

[1]: https://www.glean.com/

I feel like its very easy for companies to say no to paying for something they already live without. Unless you can prove it increases productivity.

That’s fair. I was offering Buzee for free! The inertia is real. As much as they appreciated the speed, it seemed like they didn’t really want to find their files faster.

I wonder if you could have found success asking those companies if there was anything that the tool could do in addition to helping find files that would make it worth paying for, did you by chance try asking this? Sometimes startups pivot to nearby goals, and it works out. I have a feeling out there is a company you could have appealed to, maybe not 100% with just "find files fast" but maybe "find all x files, from x date, and x client"

Everything plus massively long, descriptive filenames are the sweet spot that I'll never leave.

Want tags? Just stick the tag in the filename. Now it's a tag!

Everything has text search. Either through advanced search menu or through content:”text” - though not an indexed search. Or did you mean OCR text search on images?

Kiddan paaji? "Everything" has been my go to as well for the last 8 years!

Thanks to make this open source! Doesn't seem to have any AI enabled search feature does it? Honestly I think you gave up too early, there are solid foundations in the app but definitely needs some more polishing for practical use. I managed to build it on my mac. npm install, npm run tauri build. What is your workflow to add features / debug?

I feel I gave up early too but oh well, I sat on it for too long. I hope somebody else can make something good out of it!

Once I had a more reliable foundation, I planned on adding a tiny LLM that can run locally and parse user’s natural queries in a manner that the app can understand.

I didn’t really have a workflow to do anything since I am not a professional developer. I would just hack around in a dev branch, merge it with main and create a build when I was happy with it.

You can probably get a long way by just adding vector search without llm (if you haven't done this already)

I use recoll: https://www.recoll.org/

Recoll will index an MS-Word document stored as an attachment to an e-mail message inside a Thunderbird folder archived in a Zip file (and more… ). It will also help you search for it with a friendly and powerful interface, and let you open a copy of a PDF at the right page with two clicks. There is little that will remain hidden on your disk.

Less powerful but still pretty useful is VoidTool’s Everything

Won’t let me download? My work environment could really use something like this. Our electronic records are so vast that Windows search no longer works. And they refuse to delete anything. With paper files, after a certain number of years, we knew what to trash. That mentality hasn’t extended to electronic yet. As soon as someone gets nailed in audit discovery, I suppose…

Hi, I am sure Buzee could be of great help to you! It certainly helps me out with my work which is quite similar to your scenario.

For Windows, you should try this link [1]. Unfortunately, it'll only work if you install it under admin privileges since I couldn't afford to buy a $500 code signing certificate. Otherwise, you could build the app on your machine from the source code! I'd be happy to help you with that.

Let me know how it works out.

[1]: https://buzee.co/download/#buzee-for-windows

Ugh, I tried looking into the state of codesigning foss apps for windows, and it's not great it seems.

https://github.com/gerardog/gsudo/issues/1#issuecomment-2111...

suggests either:

https://github.com/Azure/trusted-signing-action

Which is 10 USD/month - or publishing in the windows store:

https://learn.microsoft.com/en-us/windows/apps/publish/?tabs...

https://github.com/marketplace/actions/microsoft-store-publi...

Not free either, but one time fee:

https://github.com/microsoft/store-submission/issues/15

Azure looks good, but won’t approve you unless your company is 3 years old. They include this little important detail way to late in their docs (after you pay)

Yeah, it’s weird on Windows.

I paid Apple $99/year for code signing on Mac but at least it worked seamlessly. On Windows it seemed like the code signing process varies depending on where you buy the certificate from!

I got a digicert 1y with cloud signing on azure for less than $150. There are lots of resellers, and the one I found supports individuals. Now, MS should support trusted code signing ($10/mo) for individuals.

But yes, it absolutely sucks and it’s very easy to lose hope when developing for desktop. Every platform sucks in unique ways.

Hit me up if you wanna chat. /A fellow tauri app developer

GDPR helped a lot with that in my previous companies. They applied vicious "delete after X years" policies to everything (files, emails). There is an ability to change it to longer but each element requires action, like moving to special folder with different retention.

Absolutely bonkers, as the only metric is creation time, so good luck if you have a precious folder you forgot, with old contracts or something like that.

I refuse to use a program named after a Labrador Retriever without seeing a picture of said Labrador Retriever.

Here's mine in her thieving puppy phase. https://imgur.com/a/wpJD4SI

Haha very cute!

You can see Buzo’s mugshot here: https://buzo.tools/

:D

Thank you for delivering!

Based on your explanation and my very very limited knowledge of Microsoft SharePoint, sounds you were trying to implement SharePoint!

And there's a lot of enterprises that use that and pay a lot of money for it! The business side problem i can see is : how do you convince MS focused companies to use your product instead of SharePoint? Could your product be built on top of it?

Can you for example also enhance gSuite?

Thank you for open sourcing this, so others may learn from you or build upon your work!

File search is only one piece of the massive puzzle that is SharePoint. A very important one, though!

The first link, "Download for Windows Download for Mac", seem to be a broken link to the Releases page. But the real Releases page also doesn't have any releases. I also think that a project that has the GUI as its main value add should probably include a screenshot ;)

Haha you are right! The releases were actually hosted on a separate public repository to handle automatic app updates since the source code was private.

I have updated the readme with screenshots. You can visit the website [1] or the releases page [2] to try it out.

[1]: https://buzee.co/

[2]: https://github.com/gsidhu/buzee-releases/releases

if you need to do really complicated file search, duckdb has an extension that lets you run SQL on your file system could be an option.

https://duckdb.org/community_extensions/extensions/hostfs.ht...

Thanks! Search wasn’t the most difficult problem here. I found SQLite’s FTS5 to be quite powerful. Later combined with Tantivy, I think it’s a superb combination to tackle pretty much any local search needs.

What issues did you have with FTS5 that made you switch to Tantivy? Does using Tantivy mean you have to load the entire Sqlite db into memory?

I ask because in my own project I have my own search query grammar [0] written in Lezer that more or less "transpiles" [1] to SQL+FTS5 that's been uh, tedious to build.

    [0] https://github.com/AlexErrant/Pentive/blob/main/shared-dom/src/language/query.grammar
    [1] https://github.com/AlexErrant/Pentive/blob/main/shared-dom/src/language/query2sql.ts

What were the most difficult problems you faced, if it wasn’t search?

Getting everything to work reliably. This was my first time handling a software with so many moving parts. If I could get the search query to work, I would break the date range filter, if that worked, the document indexing flow would stutter, or I’d discover a new bug caused by an edge-case.

Since I built this application in an exploratory manner (branching out without a larger architectural design), it often became unwieldy for a solo hobbyist like myself.

Plus, I wasn’t able to really get going with user feedback. Reached out to family, friends and colleagues but faced the Windows roadblock and got demotivated. In the end, built it for myself and used it by myself. I am happy with that!

Duckdb can't even hanlde edge cases in a csv (pandas does it no problem) so I wouldn't trust it to do anything useful at the filesystem level

Thank you for sharing your Buzee with the world.

(For your next project, may I suggest double checking if the name overlaps with any colorful slang?)

I named Buzee after my pet dog. I guess I am unaware of what you’re referring to, English isn’t my native language after all. Hope it isn’t a slur or something!

Made an account just to expand your Hungarian knowledge :). Around here "buzi" is a slur for "gay".

It might be somewhat localized. English is my first language and I’ve never heard this slang.

Makes sense!

I’d presume they’re referring to boozy (like booze/liquor) but not sure if that’d qualify as “colorful slang”

None of the people I dm'd didn't notice it :( a buzee packed full of docs is likely a cautionary tale about the costs of homogeneous tech communities.

https://www.urbandictionary.com/define.php?term=bussy

Code's great though. Love the trend of svelte used in desktop-focused apps.

There's an offensive term in UD for every vaguely pronounceable thing - that's not even spelt the same.

What's the overlap? (I asked Claude and ChatGPT and neither of them could figure it out.)

Omg hi Simon! Big fan. Your work on datasette and generally using SQLite for solving problems was one of my main motivations behind Buzee. Thank you for all your work and for documenting it all!

'Buzee' pronounced as 'boozey' (from the readme) is a similar to a common slang term for intoxicated or alcohol-related.

I assume it’s intentional.

Thanks for making it open-source!

btw had a request, I saw your repo lacks proper documentation. We are building an open source ai documentation system, I would love to try if I may have your permission to build docs and create a PR!

You can checkout our project here: https://git.new/akiradocs

I dream of a file system that I never have to organize. I just say what I want.

You tried to make a file search utility a startup? Did you sell any copies?

Looks really cool. Maybe the model is “your desktop for free, add other providers like GDrive / O365 for money” or maybe you tried that already?

Yes. That is the standard business model to follow. But I never reached that point because of my novice programming skills and the fact that this was a hobby project.

My intention was to make it free for individuals (search on desktop + personal cloud) but charge business users who would want shared access on a network. Any AI features would also be free if it runs locally or chargeable for Claude/GPT. I wanted to leverage Mac and Windows’ built-in AI/ML capabilities as much as possible. The OCR on both systems is actually really powerful! Even my ten year old Asus laptop was able to OCR PDFs and images very quickly and reliably.

Local and privacy first was my core principle.

kudos, builds a top tier tech stack local search interface and calls self a novice programmer .

You make my day. Thank you. This sort of validation makes my inner child really happy. :)

What @mentalgear said, amazing project!

Anyone have suggestions for an app to search across all my stuff? That is, Google Drive, GitHub, GitLab, Box.com and local files?

Sounds like an amazing idea but I installed the windows app from the website and it crashes immediately after opening.

Did you download from the link on this page? [1] That should download the v0.1.1 build. So long as you install it with admin privileges and ignore the security warnings, it shouldn't have an issue.

In case it doesn't work still, I'd suggest trying to build from source. I wouldn't be surprised if your machine is an edge case for Windows that I missed.

[1]: https://buzee.co/download/

I get a lot of this with spotlight/mdfind and it's realtime. Trying to understand if I'm missing a value add.

Spotlight didn’t quite work for me for file search. But yes, there is not much value add here other than a GUI.

The main motivation was that Windows Search is often slow and terrible. But then I wasn’t able to help many Windows users because of corporate security on their work machines and my lack of $500 Windows code signing certificates.

The main sell was that it was fast and did full text search using natural language.

I had several other plans but wasn’t really able to make much of it happen. In the end this became a passion project that I tweaked endlessly to learn new technologies rather than create value for the user.

Dope, I like the use of Tauri!

Why not Alfred or Everything?

Of course, Alfred and Everything are more mature products. My motivation was: (1) Build something on my own (2) Build a dedicated full-text search only product

I was trying to scratch my own itch really.

To this list I'll add Thomas Tempelmann's excellent Find Any File for Mac (NMNA):

https://findanyfile.app

Still, I like Buzee's UX and 10/10 agree it would be a major step up for Windows users.

I just read “I like Buzee’s UX” and giggled. Thank you for the kind words, stranger!

Alfred is amazing, but it's Mac only.

Why did you focus on Mac when they already have something that is at least presentable with spotlight/mdfind, whereas Windows search is (and always has been) atrocious? Especially in a business context. I see a much bigger market opportunity for a good search tool that is tailored to Windows.

You are 100% correct. I focused on Mac because my personal machine is a MacBook :)

That was part of my failure. I wanted to solve a problem for Windows but didn't have a decent Windows machine myself! Had to rely on a 10-year old laptop to do testing and building, which inevitably didn't work well.

Windows would also have let you target business people with specific niche business problems. You could be the best search for legal documents, invoices, etc.

Funny you say that. Legal and invoice were actually the two avenues I actually explored! There definitely is an opportunity there but I found that there is an intense status quo to fight.

Was running Windows in a VM not an option?

I tried Parallels with like three different accounts :)

But I also found that a native Windows machine throws a lot more errors than a VM. For example, having a sqlite3.dll ended up being critical on the Windows 10 machine but not on the Parallels VM. Not sure why!

SQLite is actually bundled with Windows as winsqlite. You can install your own copy if you want to take control of the version, but you usually don’t need to.

Windows users never pay for boutique apps. Only huge enterprise software or games.

Maybe now you should try to do an app to forget files. Mindfileness.

I'm sorry, but it looks like you tried to reinvent explorer/finder without adding any other useful features. You need to add something besides regular features if you want people to want to pay for it.

"did you try slapping AI on it?"

Used the operating system’s native OCR APIs for parsing text from PDFs and images. Does that count? :)

That’s the next step. Today I just wanted to get it out in the public.

Will add an architectural diagram to the readme and in-line comments to the code.

Or you could work it out and add documentation. The beauty of open source.

I think your wait list signup is not working, I put my email in but clicking submit just makes the box shake with no success message.

That's pretty cool! What you're building is kind of what my vision with Buzee was. Wishing you good luck!