Meet the Developer: Avichal Garg of Spool

The very first review that I wrote for Web.AppStorm was of Spool, an up-and-coming service now in public beta that allows you to save articles, videos, photo galleries, and just about anything else, so you can look at them when you have the time to properly enjoy it. I was very happy with Spool, and was lucky enough to get to speak to one of the founders, Avichal Garg.

Avichal spoke to me about how Spool got started, how excellent the Spool team is, and using the product that he and his team have created. Keep reading to see our conversation and learn more about the work that goes into one of our favorite new web apps.

The first question that I’d like to ask is about your background. How did you find yourself building Spool, and what were you doing before you created this new product?

Some background on Curtis and myself:
Curtis and I have been friends since I was a freshman at Stanford (he was a junior). We ended up in the same dorm and became friends. We both did Computer Science undergrads. He did a masters in CS, I did a masters in Management Science. Curtis started an enterprise software business which is still successful, profitable, and growing. He then started a semantic search company, CruxLux, which was acquired by Kosmix/Walmart. I left Google in 2007 to start PrepMe, an online education company that uses the same sorts of artificial intelligence techniques that Google uses, but PrepMe applied them to create personalized learning experiences online. PrepMe was acquired last year. So both Curtis and I found ourselves having started and sold a company, and looking for a new challenge. Since we were old friends, we were hanging out anyway and just kicking around ideas for fun. Silicon Valley is almost like grad school…everyone is immersed in ideas, and constantly analyzing, discussing, and debating which of these ideas have merit.

The people in the picture are: top, left to right: Aditya Koolwal, Curtis Spencer, Avichal Garg bottom, left to right: Ana Diaz, Dan Witte, Christine Tieu not pictured: Chandra Patni

The people in the picture are: top, left to right: Aditya Koolwal, Curtis Spencer, Avichal Garg bottom, left to right: Ana Diaz, Dan Witte, Christine Tieu not pictured: Chandra Patni

How Spool Started:

We were traveling around the Bay Area regularly for beers and coffee and we were constantly running into a very basic problem – getting to web content was far harder than it should have been. There was no guarantee that I could get on a train and have a 3g signal while I’m on that train. Or if I were in San Francisco, waiting for a friend to show up to coffee, there was no guarantee that I would be able to just pull up a short technical video I’d seen on a blog. And we realized that the world had changed underneath us. Internet content consumption is now device centric and the underlying wireless infrastructure we have can’t keep up with demand.

What we wanted for ourselves was a simple way to have our favorite content always available, without worrying about which device I’m on (my Android phone vs. my iPad), where I am (inside, outside, home, work), or what kind of media it is (text, pdf, video). And if we could build this for ourselves, we’d always be able to consume content without any friction. So we built Spool. With one click you can save content from any of your devices, and that content shows up on all of your other devices too, is available offline, and is converted into a format that will work for you. So you don’t need to worry about Flash and you don’t need to worry about whether your phone has a PDF reader. The content just shows up and you don’t have to worry about any of that friction.

I’d really like to talk about the SpoolBot for a little bit. It seems like most read/view later services have a certain script that grabs content and then pushes it to your device; what makes SpoolBot different? How long did it take to build it?

SpoolBot is an artificial intelligence and computer vision engine. SpoolBot is far more robust and sophisticated than a simple script that parses a page. SpoolBot runs on a server that we control and interacts with the actual page that you wanted. It executes Javascript, Flash, can install plugins, has a virtual video card and audio card, can access the file system on the server, and can actually see the page itself. So instead of just looking at text, SpoolBot literally sees that there’s a flash object on the page with a triangular looking thing on it and it moves a mouse pointer there and clicks on the triangle. And then if a video starts playing it sees that data streaming over the video card it controls and it grabs that data and saves it for you. Or it can see that somewhere on the page there is something that says “Page 1 of 5” and know that it should go get the next 4 pages of the article for you.

SpoolBot, doing it's thang.

SpoolBot, doing it's thang.

This makes SpoolBot very very robust. If a site changes, SpoolBot still does the right thing. Or if someone Spools a site we’ve never seen before, SpoolBot does the right thing because it’s seen hundreds of thousands of other sites and learned from all of those sites what users tend to care about on a page. Or if it’s in another language, we still figure out the right thing to do.

SpoolBot is one of the most sophisticated web crawlers out there and we have several patents pending for the technology. The Spool team has 2 PhDs and 4 masters degrees in artificial intelligence between us.

We’ve been working on SpoolBot for almost a year. It’s an on-going project and as we get more data and more users, SpoolBot will get smarter.

That’s very interesting, and I must say that SpoolBot, to someone like myself, seems like technomagical engineering.

I want to get a little bit more into how things work for you. What’s a typical day in your life like? Do you have a set process for how you work, or any favorite tools?

There is certainly a lot of magic going on. We’re lucky the team is full of such amazing engineers. 🙂

A typical day in my life looks something like:

  • 8am – wake up and check email
  • 9am – take train to SF and work on the train
  • 10am to 1pm – product related work
  • 1–3pm – meetings and email
  • 3–7pm – product related work
  • 7–8pm – dinner (usually over a meeting or email)
  • 8–12pm – email, project management, strategy work
  • 12–1am – take the train home and use Spool on the train

I try to have meetings in the early afternoon (when I’m tired) or over meals. The hard work like product and strategy work I try to do when I have lots of mental energy which tends to be mid morning and early evening.

I live on email and Spool. One of the more useful tools that I don’t think a lot of people use is Google Voice. My text messages are forwarded to email, my voicemail is transcribed, I can make calls from my computer. It’s tremendously useful.

I’m actually about ready to wrap things up, but I want to ask you about Spool one last time. Is there anything that you’re excited about that you might be willing to share with us? I love the app and have enjoyed watching it grow, and I hope that you guys can keep up your excellent work and continue surprising all of us.

The main thing that gets us excited is the scale of the mobile ecosystem. It’s amazing to us that there are people around the world using this product that we built and so many more are signing up each day. We think the scale of impact that is possible with what we’ve built is tremendous and that’s what gets us really excited. We want to have our product in the hands of millions of people and are excited to see Spool on a trajectory where that is a very real possibility.

Thanks, Avichal!

We’d like to extend a special Thank You! for taking the time to do this interview with us, and wish the Spool team the best in their endeavors to make the best way to save the web for easy viewing later. If you haven’t had a chance to try our Spool yet, be sure to give it a try. It’s free, and likely way more advanced than any read it later or bookmarking app you’ve used in the past.