Head in the clouds
- In an increasingly globalised, decentralised world, data processing companies like CloudFactory are changing the 9-to-5 work culture
Dec 3, 2018-
At some point or the other, everyone encounters a Captcha on the internet. This little annoyance, designed to differentiate between humans and robots in a reverse Turing test, asks users to input text from garbled images or to identify which pictures includes images of cars or street signs or pavements. But captchas also have another purpose—they teach machines to differentiate between various inputs as identified by humans. This is called machine learning and it is one of the ways in which automated system ‘learn’. And this is where companies like the Nepal-based tech company CloudFactory come in.
As any system equipped with Artificial Intelligence (AI) requires a millions of not just processed, but also classified or clustered data points, CloudFactory works to feed data and then label them, in other words, they ‘process’ data.
“Data processing takes a lot of effort,” says Kailash Badu, associate director of products at CloudFactory. “Currently, we have over 2,700 employees working for us.”
These 2,700 workers are not all housed at CloudFactory’s offices in Bhaisepati. In fact, most of them are college students who work from home; their offices only have 250 workers. CloudFactory hires part-time workers to input, process and label data points from client companies, most of whom choose to remain anonymous until the project is completed. The company reaches out to freelance workers who fit the requirements, based on their skills, experience and expertise. The employees can then work on the projects according to their own schedules, creating a flexible work space.
“We are often mistaken for a tech company that collects data for clients,” says Badu. “But that’s not true. Our clients collect the data themselves. We just help them process that data.”
CloudFactory is one of those companies taking advantage of the new work spaces created by the internet, namely, the cloud. With the ‘cloud’, vast amounts of data can be stored digitally and can be accessed by workers from any part of the world, without the need to come together physically. This creates freelance work opportunities for those who might not have the time to work full-time. Since large companies, especially multi-nationals, have a lot of work that does not require a lot of effort, but needs a large workforce, companies like CloudFactory provide distributed labour services like micro-tasking, data tagging and labeling.
To explain what exactly it is that CloudFactory does, Badu provides an example of a company that manufactures driverless cars. Because all of its processes are automated, the company requires a lot of data, like road signs, bridges, other vehicles, houses, footpaths, etc. CloudFactory will label and input all that data into a machine-learning system that will now be able to identify automatically what comes before the driverless car, because it has been accurately identified in a way computers are not able to do by themselves.
The idea for CloudFactory was born in 2008, when Canadian Mark Sears and his wife saw the potential for a distributed labour company in Nepal. While vacationing in Nepal, they met many young people interested in information technology and with software skills, but no tech companies to work for. Sears saw that a big data company set in Nepal could provide young Nepalis with myriad employment opportunities. Two years later, in 2010, CloudFactory started operations.
Since then, CloudFactory has provided services to companies like Microsoft; drive.ai, which works with building artificial intelligence for driver-less cars; and EmbarkTrucks. Apart from Nepal, CloudFactory currently operates in three different countries—Kenya, the United States and the United Kingdom. Employees from these countries are assigned user credentials for the platform where they actually work. The work, however, is not centralised but distributed. That means users can access data from anywhere as long as they have a computer, an internet connection and credentials to access the platform. The company assures complete data security to their clients as employees can only access the data that is allocated to them. After the completion of their tasks, the sub-systems are accumulated, aggregated and sent back to the client. CloudFactory also assures complete data security.
“We all have our own spaces. Just like with Gmail, I cannot check everyone’s emails,” says Badu.
Employees mostly work with an object-oriented general purpose programming language called Ruby, in addition to Go. For CloudFactory, a challenge has been their use of new-age programming languages like these, since Nepal’s programming curriculum is primarily limited to Python, an older language. Still, for young people looking to learn a new programming language, the company provides ample opportunities.
“We use Amazon’s server but that should not be the central point of the conversation,” says Badu. Because when you work with a cloud, it does not matter where the server is. After all, everything operates on the internet. The cloud might seem like a vague idea, one that is only just seeing application in a country like Nepal. But for a lot of the developed world, the cloud is where most of the work is currently happening, whether employees are physically present at their workspaces or not. The internet has made things simpler and is already changing working habits and patterns, like the 9-to-5 work day. In a globalised world where so much of the work is now decentralised, companies like CloudFactory are the future.
Published: 03-12-2018 08:15