In Part 1 of this series, I described how I came up with the idea for TripLingo. In Part 2, I covered my struggles trying to get funding and find a team. At the end of Part 2, I explained how I was admitted to StartAtlanta at the very last minute, and arrived just in time to contemplate whether to pitch my idea or not.
I’d just arrived at StartAtlanta and pitches were set to begin soon. For a second, I debated whether or not to pitch TripLingo. My intention in attending StartAtlanta wasn’t to build TripLingo; I just wanted to meet some engineers and try to find people that would be interested in helping. But then I realized that I had worked hard to get here and wasn’t being ambitious enough. I entered TripLingo in the pitch queue.
My 60-second pitch flew by, before I knew it they were announcing the 20 pitches that made it to the next round. I was somewhat shocked to look up and see that TripLingo had received the most votes out of all the pitches! “Oh crap,” I realized, “this is really going to happen.”
After another pitch and round of voting, TripLingo was announced as one of the 12 ideas. The 12 presenters moved to different corners of the room and the next 10 minutes was a free-for-all talent-grab. I stood there answering the questions I could and trying to seem friendly. “Please join me,” said my eyes.
After a few minutes, 8 people had gathered around. On the way out the door, I spotted the man I would come to know as “Pratik”. He was eyeing me oddly, like he might be interested but also had other options. I asked if he wanted to join. He said, “What do you need?” I had no idea what I needed, but somehow I sensed that he would be an awesome addition and was proved to be right. Sputtering for a moment, I finally blurted out “Database!” Apparently that was the right answer and now we had a team of 10.
After a bit of roaming around the building, we settled on the aptly named “Magic Room.”
The lovely and aptly-named “Magic” room.
After some personal introductions, I got up and started talking through the idea in more detail. Questions started fast and furiously. “How exactly did the algorithm work?” “What were my specific UI ideas?” “What were the use cases?”
Next, we discussed our various qualifications, divided up into roles, and clarified what exactly we wanted to produce by the end of the weekend. Our plan was ambitious, to say the least. By the end of the weekend, we wanted a functioning prototype of both the web and iPhone app.
A crucial element of our ability to be ambitious and yet another area where I got extremely lucky was the breadth and complementary skills of our team. We had a designer, back-end specialist, Ruby specialist, mobile back-end guru, mobile front-end guru, and marketing talent. I can’t emphasize enough how lucky this was. Our team composition allowed us to break our work into different modules and meant that there was never a point where we had people sitting around with nothing to do.
By the end of the day, we’d assigned responsibilities, had a vague idea of what were going to accomplish and agreed on what needed to move forward. A little after midnight we dispersed and agreed to be back in the Magic Room by 10am the next morning.
StartAtlanta Day 2: Ups and Downs
First to arrive Saturday morning, circa 7am!
One of the things I had to have ready by the time the team arrived on Day 2 was a complete lists of requirements in terms of the database. Luckily I’d made a prototype of the database in Excel, which gave me something to work from. But it needed a good bit of work to be ready, so the next morning I was the first to arrive and got to work laying out the database. Ross arrived at 8am to attend a workshop to set up the server and database. The rest of us trickled in and we were off to the races.
Fours hours later, and we were starting to show some signs of progress. However, Ross was having some serious issues setting up the server. Our ambitious project was on a tight schedule, but as the clock ticked by we still didn’t have a database or server. This delay was holding back everyone and the level of disquiet in the room was gradually becoming more visible. By 4:00pm frustrations were being vocalized. A core piece of our architecture simply wasn’t coming together. Maybe we should roll back our scope? Were we even going to have anything to show for it?
I was becoming increasingly tense. I was trying to keep everyone happy by bringing drinks and snacks to the room and make vaguely optimistic statements. But for practical purposes, I could only watch helplessly as our morale gradually deteriorated. We’d hit an obstacle. To our team’s credit, we kept working, kept hacking, kept trying. A suggestion that “maybe we should throw up a white flag” was quickly shot down. We were not going quit? We were not going to quit. We were not going to quit!
A little before 6pm, something miraculous happened. We figured out what was wrong and at long last we had a functioning database! Hallelujah! Essentially, the root of our problem was that one of the versions of database we were using had a weird bug. Our mood lifted immediately, and suddenly we were back on track.
We didn’t stop working that day until 3am. We still had a lot to accomplish, but there was light at the end of the tunnel. After the last person left, I undressed and snuggled up in the sleeping bag I’d brought. Day 3 was approaching fast.
StartAtlanta Day 3: A Race to the Finish
The next morning, I was awoken to the sound of Brendellya saying “Jess, wake up. Wake up!” I thought I was dreaming. I remember wondering why the heck Brendellya was in my kitchen? When she said “Its 9:00am”, I finally erupted out of my sleep, got dressed, and got moving.
Day 3 flew by. We knew from the beginning that the ambitious nature of our project might have us cutting it close to the 5pm deadline, but had we known how right we were we would have reconsidered.
Work work work. Hack hack hack. Deadline looming!
In any case, the next 8 hours was a frenzied hackfest. I spent much of the day crafting and practicing the pitch. The rest of the team worked nonstop to get the prototype functioning. Shortcuts were taken, hackers hacked, and with just 15 minutes to spare we finally had the prototypes we’d been working for.
At 6pm, we assembled in the main conference room. By 6:45 the pitches had begun, and around 7:30 it was my turn. Rather than flying by, the presentation seemed to take place in slow motion. Watching the video, there are times when I forgot a line or a transition and I remember in my head trying to improvise and considering different ways to move forward.
The final pitch at StartAtlanta.
In any case, after 3 minutes of presenting and 5 minutes of questions, it was done. There was nothing to do but sit back, enjoy the other presentations, and cross our fingers.
After all the presentations, it was time for the voting. Attendees handed their nametags to the team representative they wanted to vote for and then the votes were tallied. After not much adieu, the results were announced; TripLingo had been voted the winner!
A round of high-fives and congratulations was in order and the next 20 minutes was an adrenaline-filled period. Needless to say, I was thrilled our team had won!
But there’s a reason this story falls under the “birth” of TripLingo; this was just the beginning, and the real work lay ahead.