Ask YC: Are there any promising non-web startups?

[mikelikespie]

I'm graduating soon and the job search has began. Unfortunately, it seems like there's not a whole lot of opportunities for somebody like me in companies that aren't huge corporations. I can do javascript, and web applications. I find doing C, C++, and python much more fun, especially when dealing with concurrency, optimization, 2D/3D graphics, image manipulation, etc.

So I guess my question is: Is there a place in the start-up world for programmers like myself? or should I just suck it up and sell myself to a large corporation or spend my time concentrating on becoming proficient at web stuff?

[ivankirigin]

Some smart robotics startups:

MekaRobotics (by Aaron Edsinger, MIT star behind Domo) http://mekabot.com/ http://people.csail.mit.edu/edsinger/domo.htm

Willow Garage (former early Google employee founder) http://www.willowgarage.com/

AnyBots (Trevor of YC's company) http://anybots.com/

Q Robotics (early stars from iRobot that liked startups so much, they needed to leave and start another once iRobot went public) http://qrobotics.com/

You don't really need to know anything about robots to work on them. The best groups are made up of specialists that understand a bit of everything, but are essentially software|hardware|electrical|systems engineers at their core.

[bayareaguy]

One thing you could do is follow the VC money. Do some VC research. Look for VCs that invest in whatever non-web stuff you think makes sense (you must have some ideas there). Then look into what startups they've funded and consider contacting them or their recruiters.

If you're more ambitious you could research the kinds of problems you think those startups will have and look for an application you can build that they might need but isn't something they would want to develop themselves.

[emmett]

The web is just the most convenient place to deliver software.

Justin.tv has a huge* amount of highly concurrent, high performance Python code that delivers our video. That video is still delivered via the web, though.

Speaking of which, we're hiring...

* It's not actually that many lines of code, but it's comparable to our web codebase.

[justin]

Thanks for pointing that out, Emmett, we are indeed hiring! Both our chat server and Video cluster are written in Python's Twisted framework. We're always looking for smart hackers who want to help us make our systems better. Everyone at JTV is tossed in the deep end (Bill, who found us through news.yc, learned Twisted, wrote and deployed our chat server in his first two weeks here). Find out more here: http://blog.justin.tv/2007/07/justintv-is-hiring.html

Or email justin at justin.tv

[manvsmachine]

I would think that there are a decent number of non-web tech startups, but not as many straight-up software startups. Oftentimes, small non-web software companies will get acquired as soon as they come out with something useful, and their ideas are integrated into the larger companies' products. My suggestion would be to do the big company thing for a few years, and then once you have the experience/contacts, you'll be in more of a position to find a fledgling startup or form your own.

[cperciva]

Not all internet startups are web startups. In my case, I'm working on online backup, and every single line of code I've written for it has been in C.

I'm not going to suggest that you should get into the same field and start competing with me, but there are many options under the umbrella of "infrastructure services" which all require highly efficient coding.

[mrtron]

Certainly. Although most internet startups will include a strong web presence, that is often a small part of the equation. I got pretty heavily flamed for suggesting to not write your own framework, but I think the whole point was different business genres. For an internet startup, you probably don't want to write your own framework. For a web based startup, where that is the core of your product, you might have a huge competitive advantage writing your own.

Best of luck!

[neilc]

There is lots of promising non-web startups -- the fact that this question even needs to be asked is evidence of how insular the YC news community can be at times.

The field I know best is database systems -- there are a number of interesting startups in this area. There's an emerging market for "complex event processing" (essentially database systems that operate on live streams of data as well as static historical data), which has applications to algorithmic trading, network monitoring, real-time business intelligence, military applications, etc. I work for one startup in this field (http://www.truviso.com -- we're hiring!), but there are several others (Corel8, StreamBase, Aleri, etc.)

There's also data warehousing, in which there are a lot of startups challenging the existing players: GreenPlum, ParAccel, Vertica, DATAllegro, C2 Appliance, etc.

More broadly, I think non-web startups are typically trying to solve harder and more interesting problems than the web guys. The time-to-market is often longer than with a typical consumer-oriented web startup, and if you're trying to sell to enterprises, that brings a very different set of sales challenges than selling ads on a website, for example. That often means you need a founding team with a deep technical background, and you've often got to take the traditional VC route (rather than bootstrapping). So I can understand why it's not a good fit for the traditional YC model. Still, there are a ton of startups out there, and many (most?) of them aren't developing web apps.

[igexome]

I'm in the same situation - but its more of a psychological hurdle to apply to places that want X years of experience, when all I have under my belt as a soon-to-be new college grad is summer internship experience. Just be vigilant on craigslist and sites like SimplyHired and Indeed. There are tons of opportunities available and your net may simply have to be cast wider.

Oh, also - check out http://www.ventureloop.com for culling startups. I've found that these small companies have a ridiculously fast callback rate.

[iamelgringo]

If you're thinking about C code, you're probably looking at drivers or unix programming. I'm sure there's database companies that are looking for decent C or C++ programmers. You'd be programming for concurrency and optimization there. Granted, databases aren't really a sexy field, but... where there's muck there's brass as they say.

From what I understand, a lot of the C++ programming is windows stuff. But a lot of the Windows programming at small companies is going over to C# stuff. But, there are some ISV's around that are probably looking for programmers. Joel Spolsky is always hiring. I don't know how much concurrency work you'll be doing at an ISV, though.

As far as 2D/3D graphics are concerned, there's always work for programmers in the special FX industry. Those tend to be smaller companies, and they tend to do more C++, C, and python stuff working with graphics. You don't have the upside potential that you do with startups, though. If you're interested, by a copy of CineFX at Borders, go through an look up the company names and see if they're hiring programmers. Or, Goiogle companies like Pixar, Dreamworks, ILM, Tippet studios, Blue Sky and see if they're hiring programmers.

The VFX world does a lot of C++, Python and works a lot with distribued computing, render farms, graphics, optimization, image manipulation, etc... It might be your cup of tea.

[henning]

No. The only startups that exist are JavaScript hacks that take up 90 MB of RAM.

[hhm]

Didn't you think in applying for research in places like http://www.merl.com/, http://groups.csail.mit.edu/vision/, or in companies such as http://www.gesturetekhealth.com/, http://www.tandentvision.com/, etc? There is lot of interesting research in computer vision in usa, both in large and in small companies and companies and laboratories if you are interested in that. Maybe some of these companies are small enough to categorize as "startups", I don't know. You could check the always growing robotic field too.

This site could be useful for you, too: http://www.cs.ubc.ca/spider/lowe/vision.html

[falsestprophet]

There are a lot of cool computationally intensive problems in drug design/discovery and trading/investing. There is also the possibility of a substantial upside in both fields.

[DaniFong]

We're starting a startup in computer vision. We're deploying across the web, but that's just because it's the cheapest distribution mechanism.

You describe interests that might imply you'd enjoy game development. Have you thought about giving that a go?

[mikelikespie]

I think computer vision is cooler than gaming personally.

Actually, my skills are perfect for game development, but it seems if you want to work as a game developer you have to sacrifice pay or something else just so you can work on games. And the fact that I would work on games wouldn't really compensate for the sacrifices because I'm not a hardcore gamer that went into computer science just to write games.

[asmosoinio]

Computer vision, neat. You guys need a good developer? If so, you can find my email in my profile.

[tx]

http://www.neuric.com

These guys are in Austin and they're building some cool AI technology. I have seen the demo and it was very, very impressive (their AI has emotions, so it can be pissed at you, etc). I suppose I cannot disclose much more, even though I do not work for them, but they do not use JavaScript for that.

BTW: There are other startup hubs, besides SV, in US. Austin is one of them, and I would say that most companies here do not do dot-coms, often it's hardware+software combos. Recently I bumped into another local startup that does some cool stuff with visualization - they've been hiring some C/Linux hackers (don't recall their name).

I suggest you burn a custom RSS feed for jobs@craigslist with certain keywords, like "startup", "software", etc.

[jdvolz]

Coming from the consulting angle, it is much easier to get work doing web related stuff, and that work pays better I am finding. Even the desktop stuff I am doing either has a web component, communicates with the web or a website in some way or has a version that is a website. Learn your Javascript and have a decent working knowledge of HTML and CSS.

I do think that you should know probably PHP (for consulting most web projects are in this or are extensions of projects written in it) and one of either Ruby or Python.

There are opportunities out there, just not nearly as many. Also keep in mind as a recent graduate you are competing with all the people who never made the web transition because they already had 3-5 years experience on the desktop that now have 10-12 years experience there.

[thorax]

You might check out cilk: http://www.cilk.com

[utnick]

there are tons of startup video game companies that you would probably be interested in

[mikelikespie]

Maybe I'm not good at finding startups.

[menloparkbum]

video games, hedge funds

[wmorein]

Xobni.