How SeeFood could have been a real deal in ‘Silicon Valley’

The latest season of HBO’s Silicon Valley has been one of the best comedy shows on television. Such has been the popularity of the show that it has garnered fans from the Valley and beyond. Along with good drama and comedy, there have also been some ingenious product ideas that one can ever think of. No, we are not talking about the 8 recipes of Octopus app using Jian Yang’s grandmother’s secret recipe. This app is beyond genius and should not be discussed because Jian Yang’s culture won’t feel good about it.

While Richard was trying to figure out his idea of new internet using his data compression algorithm, Erlich came up with this new app which would become Shazam for food and all this happened during his pitch to a VC firm, Coleman Blair. The app would be known as SeeFood — “like a food that you can see,” according to Erlich.

As soon as they left the meeting room, they knew that it’d be a difficult task to actually build SeeFood. It needed lots of data to train machine learning algorithm and all this had to be done in short duration of time. So Jian Yang decided to take one step at a time and made the Not Hotdog app, which could easily classify only hotdogs, i.e., it could tell whether the food item was hotdog or not.

Building Silicon Valley’s Hot Dog App in One Night_After watching the latest episode of HBO’s Silicon Valley and seeing the Not Hotdog iOS app that was built, I wanted to…_hackernoon.com

But, that wasn’t good enough to be called as SeeFood, as it couldn’t recognize other food items apart from hotdogs. Think of it like Shazam that would work only for one song.

As Dinesh pointed out, the core tech was valid and it just needed to be trained. So, what Jian Yang did for hotdog, now he had to do for all other food items in the world. But, Jian Yang was not inclined towards this not-so-interesting work of scraping the internet for thousands of food pictures. He suggested that they could hire someone else to do that.

But, Erlich explained that they already spent huge amount of cash on Dinesh to help them and there were other expenses like legal, marketing, operating fees, AWS charges, etc. Thus, they were not in a position to hire people to do this manual work of web scraping for them. However, Erlich went on to build a Palapa using VC money so he could upscale his workspace to help with recruiting.

But, Jian Yang was done with this idea as he was stuck with this problem of scraping thousands of food images from the web so that he could train the app. So, he went ahead and made big business expenses by buying a Corvette for personal use which eventually left no money for SeeFood. Now, Erlich tried to finish this work using Nelson’s students at Stanford University. But, even they outsmarted him along with Coleman Blair who decided to fund them.

Now this is where using web scraping services can be a big saviour. It costs less, gets you accurate data in ready-to-use format, saves lots of your time and allows you to focus on your core application development rather than worrying about acquiring data from the web. It costs a lot more when it comes to hiring skilled labor and servers to maintain and monitor the entire process.

In this world of cutting edge technologies and data driven businesses, you need to be constantly vigilant, as information and data are key for both survival and growth. If you get the right data at the right time in an efficient manner, you can stay ahead of competition. If Erlich and Jian Yang had been smart about this, they could have quickly started a project with a web scraping service provider and acquired thousands of food images from the web in the same time instead of paying a huge chunk of money to Dinesh. The process is rather simple — identify the list of food photo sites that allow crawling along with relevant data fields. Then, let the Data as a Service provider use the existing crawler infrastructure to extract data at predefined frequency and deliver the same in structured format so that it’d be directly consumed by the database. In fact, by following this they would’ve saved enough money to build a Palapa and buy a Corvette too!

Do you agree? What are your thoughts? Let us know in the comments.

Silicon Valley airs in the US on HBO on Sunday nights.