When I was a freshman, I originally wasn't planning on getting an internship. I was just going to work on Class Checker over the summer. Then right before the summer started an email went out to the CS students about an internship at Shoeboxed called "Start your own startup". They liked the projects I had done and commented on how it was impressive for a freshman, so I got the offer. I was working with a graduate student in business at Duke and a recent grad from NC State. I was the only tech guy, but my teammates were very sweet and supportive. We played a lot of ping pong that summer, and I had a good time hanging out with them.
read moreI also learned a lot about the business side of starting a company. Every week they had a local founder come in to tell their story and we'd ask them questions. They took us to local incubators, like Triangle Startup Factory, and I got to hang out around Durham which I hadn't done. Often I'd get to work early, and spent time in a local jazz cafe eating breakfast. It was pretty idylic.
Tobi and Taylor at Shoeboxed were really amazing, to pay interns to just work on their own project that didn't necessarily have to do with their business was very generous. Ultimately, I'd say the project was a failure. We went between a couple different ideas and landed on one of mine, which we called Tagzly.
The biggest takeaway was learning the kind of culture I wanted to be around (and eventually create at Coursicle). They threw out the standard company culture: big sales were rewarded with alcohol, people would play tons of table tennis, there was once a softball game that they played inside the building (using a table tennis ball).
Here are some of the ideas we considered/started building:
- A geofence app that sends you a push notification when you're near a business with a discount. We were going to scrape all business locations from Google Maps and set geofences around them. Geofencing was a hot technology at the time, which is probably why we looked into this.
- I think they were pushing us lightly to create an app that would scan a grocery list and come up with the stores you should go to to buy the shopping list for the cheapest, but we couldn't figure out how to get that data and it didn't interest us enough.
- A toilet finder app that would tell you where the nearest public toilet was to you. We were all pretty stoked about this but the only data we could pull was state public restrooms which were just along highways, not inside cities/towns.
- Taglzy, an app that allows you to track mentions of certain things on the internet. Frankly, it was a glorified Google Alerts/Reader, with the ability for more fine tuning. The Shoeboxed people recommended using Django because I had some experience in Python, but this is the internship where I learned I hate using frameworks. I was stuck on a bug for multiple days that I needed to ask Stackoverflow for help on. As a newcomer, I also found all the boilerplate confusing. In the end in order to get it to work, Taylor (the CEO) and another experienced developer (Curtis) had to spend a couple days finishing it up. It felt pretty dissappointing since we only got a couple dozen users despite my partner's extensive marketing efforts and working on it most of the summer. I couldn't revive the website, but I did find the blog that Eric maintained while we were working on the site.
Since the name of the company evokes the idea of a literal Shoeboxed, I setup a little joke with Legos that I sent Tara.
Tagzly notes
Discover is a feed of all the tags that are trending and related to what you currently follow. My tags is simply a list of all the tags that you're following. You can drag tags from discover into the "my tags" section to start tracking. Perhaps My feed would instantly add the top article for that tag at the time it was added so that they user receives instant feedback. It could slide in on the stack of my feed. The user could scroll through discover to find more tags to follow. Discover could be an infinite scroll. Maybe you could hover over a tag in "My Tags" and then you can click and resize the tag to specify relative importance to you. You could edit a tag and customize it. You would have the ability to specify which articles you want pushed to you: you could say I want the article to have the word "apple" but not "droid" and the words "new operating system" OR "new iOS" all in sequence. As you're making the tag, there could be an indicator that estimates how many articles per day you could expect to match that tag (after we filter for importance). More relevant example: any articles that have "boston bombing" and "new" and are within 24 hours of the date I'm viewing them. How do we establish which articles are the important ones? tfidf just determines which articles are the most relevant, but what if the most relevant isn't the one they want to read? Various parameters to narrow searching by: include word, exclude word, sequence of words, subject contains, how recently was it published, time of day published, when to show it to me (in the morning, evening, etc), has a video, is by a certain author, contains a keyword in the RSS keywords (don't think all rss feeds have keywords), send me push/email/text notifications with each update. See twitter posts that have the keywords in them/have the keywords as hashtags? Allow the following of subreddits (they have rss feeds). Maybe make it so that you can make reddit comments on the articles from the website. When the user enters in tags, we do a general search with the google find api. We then take the articles it supplies (note that it supplies 10 top articles and also the feeds they were taken from) by getting the LINK from each entry and then follow the rss feed to find future articles by getting the URL from the entry. We use these top 10 results to initially provide them with the best articles, and we keep track of the feeds they came from so that we can push them content from the same source later if they want it. NOTE THAT WE MAY BE ABLE TO JUST DO THE GOOGLE FIND API FUNCTION EVERY SO OFTEN TO UPDATE WHICH ARE THE TOP 10 ENTRIES. THIS MAY BE BETTER THAN EVEN TRACKING THE FEEDS AT ALL.
Shopping list app notes
APIs that might be of use This has a lot of stuff and looks like it does specifically local stores. Free.: http://developer.goodzer.com/docs/ First one I found: https://www.semantics3.com/usecases/ecommerce Perhaps for coupons and deals: http://www.popshops.com/support/coupons-and-deals. Also http://www.popshops.com/signup Groupon is something else we could integrate.
Geolocation app notes
Idea for getting every single restaurant on google maps: What you do is start with some location, find all those around it that match a criteria, then for each of those do the same query. ad infinitum. FOR ALL: Report the Events listed for a place. Could include start time and summary and URL of event. FOR ALL: As a backup, you could display a high starred review for whatever place. FOR ALL: If you're within an hour of closing time, get a notification that a place is about to close. If you're lingering, let them know (so that they can clear out). FOR ALL: Display top rated for whatever they're selling. Airport: [holiday] is coming up soon, do you need to go see family? Amusement Park: This park seems to have especially low prices, perhaps you should join the fun. Aquarium: Starfish are not fish. Neither are jellyfish. Art_gallery: You've been working a lot recently, why not quickly browse [nearby art place] ATM: You're good on cash, right? Bakery: You just came from a restaurant, how about a pastry? Bank: Have any checks you need to deposit? Bar: Maybe there's a game you could catch at the bar? OR display top rated alcohol Beauty Salon: It's been a while since you've been; want to make an appointment [phone number]? Bicycle Store: There's nothing greener than a bicycle. OR Facts about Book Store: *Display new york times best sellers* http://developer.nytimes.com/docs/best_sellers_api Bowling Alley: Go with Eric or Justin Bus Station: DIsplay local routes (really hard). Cafe: This cafe is the cheapest around. OR this cafe is the most expensive around. Campground: How about a family camping trip this weekend? Car Dealer: [company] sells cars for the cheapest price. Car Rental: You're in a new city. Need a rental? Casino: Americans spend X amount on Casinos every year. Cemetery: Off limits. Church: Off limits. City Hall: History of this city hall? Events? Clothing store: Check out these fashion tips OR Here's what's in style OR this place is the cheapest/most expensive in area. Convenience Store: There's a place down the road with much better prices. Courthouse: Give some history of the courthouse. Dentist: Hope you've been brushing! Department Store: