Every crisis forces enterprises and governments into 3 modes: React – quickly assessing the impact of the crisis; Recover –taking short term measures to absorb the impact; and Respond – planning and executing long-term strategy. Over the last 6 months, COVID-19 has pushed enterprises into the 3 Rs modes and needing to quickly make changes to their Priorities, Processes (infrastructure), and People in order to navigate through the crisis. The circumstances of this crisis are nothing new. The same thing happened 10 years ago, when the bursting of the housing bubble pulled global economies into recession. In that recession, companies were able to quickly change priorities but not processes and people. They didn’t have companies like Amazon to provide cloud services, which in turn could help them change their processes faster. Today, while companies are coming out from React mode to Response mode, cloud has been at the core of the process transformations that have followed changes in priorities. However, the people factor is changing as slowly as it did 10 years ago.

Crowdsourcing (on-demand talent): A solution to “People” transformation

Process change and people change go hand in hand. For example, if you need to quickly scale up or scale down the infrastructure, of course AWS helps you to do that, but at the same time, you need to have the right experts to support those infrastructure changes. Today, due to the “people” element, companies are not able to fully leverage cloud capabilities. Imagine this: It takes two minutes to procure a platform or software, but companies spend on average two months to hire talent. So, ultimately, it takes two to four months to get that software running in the company; thereby, negating any value the cloud brings in. The question is how to create a “cloud version” of people or talent. Enter Crowdsourcing – Talent as a Service.

Crowdsourcing, or the “gig economy,” has changed the way people seek work. A recent McKinsey study showed that up to 162 million people, or 20%–30% of the working-age population in Europe and the United States, engage in some form of independent work. Although the likes of Uber, YouTube, Deliveroo, and Menulog have always used independent workers to provide services, companies like Topcoder, Kaggle, and Upwork have started to engage with enterprises to leverage on-demand talent for IT programs.

In the typical business setting, business managers define tasks and then try to find someone with the skills to accomplish them. That works, but it’s expensive, time-consuming, and incredibly difficult in today’s market, especially for tech roles. We at Topcoder, help businesses define a problem, and then take over the problem, break it into chunks and crowdsource each piece to develop a robust solution.

We call this “hyper-specialization” or “atomization.” Topcoder breaks a problem into small pieces and holds a competition to solve each one. Topcoder usually doesn’t use client environments. We spin up environments in AWS and invite the crowd to build the small pieces. After assembling the pieces, we invite clients to check the built solutions. Once agreeing to a solution, Topcoder will run security scans (Black Duck or Checkmarx) for malicious code or backdoor entries and then, through containers/API/GitHub code drops, push the solution into the client environment. With AWS providing more features for collaboration and an easy and secure way to move code, using crowdsourcing has become easier for enterprises.

Crowd+Cloud: The ultimate combination

Crowd and cloud is a powerful combination. Both are on-demand and provide elasticity, which is the need of the hour for companies recovering from COVID-19-related challenges. Although most companies have reset their priorities, resetting the way processes and people support these priorities is imperative. With crowd and cloud, the possibilities of reducing costs, accelerating innovation and experimenting with new ideas is endless.

Taking idea to Minimal Viable Product in days

Companies have hundreds of ideas resulting from the current crisis. The only way to move forward is to filter down the ideas into potential opportunities, quickly implement them in the scale needed, learn, and refine. This fail fast approach will help companies get off the starting line to take an early lead in the market when things return to normal. With crowd and cloud, an idea can achieve MVP in just days. We had multiple opportunities to demonstrate the power of crowd and cloud in this space and we are sharing a use case below.

Going serverless: Using AWS Lambda

One of our clients had identified Oncall and IRIS as their tools to alert teams on app/server failures. Oncall and Iris were developed on Python code base and free. However, using them as-is would require spinning up a server and incurring hosting charges.

Topcoder launched a development challenge where we invited top developers to convert IRIS and Oncall code base so that it could be packaged as functions in AWS. Leveraging AWS Lambda, we were able to make both apps serverless.

Since we were not able to invite the crowd to client servers, we spun up our own AWS instances and built the app in the following architecture.

Imaging getting this done through traditional means. With a team on prem, development would have taken up to two months to complete. Topcoder completed this project in 15 days! Now that is what we call speed.

On-demand resources

With cloud and crowd, you can have resources at your disposal on demand. Do you want to test your app with multiple mobile devices? No problem. AWS device farm takes two minutes to activate, and Topcoder can give you testing resources within 24 hours’ notice. Imagine running a test cycle by assembling infrastructure and people within 24 hours and then disbanding both immediately when you finish testing. That is a quick process and it saves you money. We are sharing the use case in detail below.

Test Android and iOS devices on AWS Device Farm

Most of the app development teams rely on manual testing for their applications. This process can be extremely time consuming and complex because of the large variety of devices on the iOS and Android markets. Testing becomes even more difficult because of multiple model variations, OS versions, carrier customizations, and more.

AWS Device Farm addresses this challenge by letting developers test their application on a shared fleet of 2500+ devices or on a private device lab in the cloud using real non-rooted devices not simulators. When paired with a crowd of 100 testers at the location of your choice, you have a truly powerful real-time device test for your app. We ran a real-world test for a Singapore bank where we invited a group of 70 testers consisting of UI/UX experts, Quality Assurance analysts, and app developers who raised 88 defects in just 2 days!

Harish Nachiraju: Harish is an Engagement Manager at Topcoder, a crowdsourcing platform that connects companies with a community of over 1.2 million of the world's best technologists. He is passionate about technology and is always attracted to avenues where technology is helping convert customer voices and insights into business ideas. Currently, he is helping customers adopt the On-Demand Talent model as a new way of solving enterprise IT problems.

Aviral Sachdeva: Aviral is the Senior Alliance Manager for Cloud Services at Wipro. In his role, he is responsible for managing partnerships with hyperscalers for the APAC region. Aviral enjoys working with cloud service providers, has a strong appetite for cloud strategy and customer experience, and ensures solutions are relevant, engaging, and exceeding customer expectations and desired business results. He believes in the power of collaboration and is keen on fostering partnerships in the area of cloud transformation.