How Artificial Intelligence is powering tech crowdsourcing

Relevant for GS & Essay:-

  • Much of today’s software product and hardware infrastructure is delivered through “as-a-service” model. “As-a-service” just means the ability to rent and pay for software and hardware as and when it is used, rather than buying an expensive licence or powerful computers. Software behemoths such as Oracle and SAP have been pivoting to this model, as have cloud computing infrastructure platforms such as Amazon Web Services (AWS) and Microsoft’s Azure.
  • The rent-versus-own decision ebbs and flows over time. Not so long ago, before India’s mobile phone revolution, the country was dotted with ISD/STD call booths in a classic example of “as-a-service” delivery. You didn’t need to own a landline phone when you could simply walk up to a booth and, for a fee, use someone else’s phone for a short while. Once it became cheap enough to acquire and use a mobile phone to make calls on one’s own, these booths disappeared.
  • Classical economic thinking would predict that this sort of “rental” model will not work for cutting-edge software application development, which fundamentally alters the workings of a firm. Firms in the quest for digital disruption will be looking to keep their competition out, and so would prefer to have their own in-house development rather than renting it. This is especially true of startup firms in the technology space.
  • Classical thinking has its limitations, however. Silicon Valley and cutting edge technology are strange beasts, and the extraordinary push into the digital space has meant that top-notch software engineers are in short supply. This has pushed up the average wage of “Google-quality engineers” to the point where their average wage tops $350,000 per year—at least so say Jonathan Siddharth and Vijay Krishnan, the founders of Turing (turing.com). Turing is itself a startup that is trying to fix this demand-supply mismatch using a “rental” or “pay as you go” model.
  • While talent is global, opportunity is not. Hiring top engineers locally in Silicon Valley is costly, and not scalable. In addition, says Siddharth, employee retention data paints a bleak picture. The average Silicon Valley engineer retains for 13 months. When time to hire, on-board, and handoff are accounted for, employers only get about nine months of productive work from each engineer. According to the duo, this is a contributing factor to one in every 10 startups failing within the first 12 months.
  • Turing wants to help startups hire pre-vetted, high-quality engineers sourced from the global talent pool. The vetting is done using automation and Artificial Intelligence (AI). Turing claims that its platform helps developers from all over the world by offering them an opportunity to participate in Silicon Valley and allows startups in the valley a fighting chance of being able to access and retain top-quality talent. Turing claims that its customers can just push a button and hire exceptional pre-vetted remote software engineers on demand.
  • In this age of open-source programming, and the ready availability of free programming libraries with a simple click of the cursor, startups and corporations are increasingly open to the idea of turning to the programming community at large, in a phenomenon called “crowdsourcing”. IT services firms are certainly wise to this phenomenon; Wipro already owns Topcoder, the world’s largest platform-based crowdsourcing service.
  • When pressed on why their platform is different from the ones that already exist, Siddharth and Krishnan say that their secret sauce lies in their ability to use AI and automation to constantly monitor the quality of programmers on their platform. Much like any other marketplace such as ride-hailing, where ongoing input from customers about the service quality of drivers is an integral component of the algorithm, Turing uses ongoing testing of its contractors’ work performance in order to continue to provide only top-quality talent, and eschews dependence on what the programmer’s resume says.
  • Siddharth and Krishnan are both Stanford postgraduates who have built a successful technology business before. Their first company, called Rover (roverapp.com), was acquired by Revcontent (now part of TechCrunch) for “north of $30 million”. Rover had used AI methods to provide targeted personalization and content, much like Facebook and Google do in order to push hyper-personalized advertising. Firms that didn’t have the in-house AI muscle to build personalization engines, but still had use for hyper-personalized messaging and content, turned to independent startups like Rover that were providing these services outside the behemoths.
  • The duo claims that the learnings from Rover, where they played on the same field as Facebook and Google and yet built a successful AI business, have made them adept at differentiating their offering from the big boys. The AI analyses data from email, actual programming code, Asana (a work management tool), Slack (a workplace communication and workflow tool) and various other communication platforms to ensure quality of work output from each developer and enforce remote management best practices natively on the platform.
  • Time will tell whether they can pull off another success despite the presence of heavyweights in their market. While we wait with bated breath this month for earnings announcements from Indian information technology majors who work on project and deal levels, the ground may just be shifting under the majors’ feet to “as-a-service” models.

Leave a Reply

Your email address will not be published. Required fields are marked *