WeShine Tech Equips Universities to Proctor Exams Remotely with Machine Learning on AWS
Artificial Intelligence for Virtual Proctoring
When lockdowns began in India in early 2020, schools needed to find ways to facilitate distance learning and testing. Pune-based
WeShine Tech—a startup focused on university examination process automation—has been working to help colleges and universities make a digital shift since 2011. The business has been introducing a range of applications over the years, but the lockdown propelled WeShine Tech to quickly adapt its flagship enterprise resource planning (ERP) UniApps product to support remote exam proctoring with artificial intelligence (AI).
UniApps is an end-to-end suite of applications that automate university examination processes, offered via an a-la-carte service model to local universities in India. The UniApps suite also provides a student lifecycle management system, from onboarding students online to issuing certificates. Some applications include a Student Enrollment System (SES), an On-Screen Marking System (OMS), an Online Examination System (OES), and a Result Processing System (RPS). More than 20 universities in India, including Savitribai Phule Pune University, use UniApps. Each university has between 300,000–700,000 students who sit for major exams twice a year.
“Since migrating to AWS, we’ve been able to deliver a seamless user experience with no interruptions.”
Anupam Mangudkar
CEO, WeShine Tech
From On-Premises to Cloud-Native Containers
WeShine Tech launched its applications as individual systems running in customers’ on-premises IT environments, at the universities’ request. But scaling became challenging as the number of universities and student users rose.
In 2017, the startup decided to run its applications as Docker containers. “We wanted to look at how we can leverage new technologies and simplify deployment. Our adoption of containers triggered the business to explore cloud technology and auto scaling mechanisms,” says Anupam Mangudkar, chief executive officer of WeShine Tech.
This led to WeShine Tech’s next move in 2019, when it chose to consolidate its applications under the UniApps suite and migrate OES within UniApps to Amazon Web Services (AWS). “We chose AWS for scalability, availability, and its overall ability to integrate cloud-native services,” says Mangudkar. As of October 2021, the business is migrating its remaining applications within UniApps to AWS, working toward a highly scalable and available cloud-native architecture.
Authenticates User Identity to Aid Remote Proctoring
Shortly after migrating OES to AWS, lockdowns in India presented a challenge for universities by coinciding with peak exam periods. An institution approached WeShine Tech for a solution to remotely proctor exams. At the time, few tried and tested solutions existed in the market, particularly ones that catered to academia. WeShine Tech took up the customer’s request and built a facial recognition service using
Amazon Rekognition within a week.
To verify the test-taker’s identity, students’ university identification (ID) cards are uploaded into the system and stored in
Amazon Simple Storage Service (Amazon S3), with personal data protected using
AWS WAF – Web Application Firewall. OES authenticates each student upon login to ensure the test-taker matches the image of the student on their ID card. It then continues capturing images of the test-taker at regular intervals throughout the exam to confirm identity. WeShine Tech also introduced a feature request to detect if a user’s browser goes out of focus, which could indicate fraudulent activity.
Ongoing Improvement with Fast Feature Delivery
WeShine Tech continues to improve OES based on customer feedback. Another feature request introduced is the integration of Zoom for video proctoring during online exams. This was done by adding a further ID validation layer using Amazon Rekognition application programming interfaces (APIs).
The ease of integrating services on AWS, and the user-friendly design of services such as Amazon Rekognition, allows WeShine Tech to deliver new features quickly and continually improve its services to boost customer satisfaction. “We’ve carried out most of the building and migration work ourselves. Integrating AWS services and building new features within the platform is logical and intuitive,” Mangudkar adds.
Handles 350,000 Concurrent Online Exams
Since migrating OES to AWS, WeShine Tech can handle 350,000 concurrent online exams daily during peak periods, up from 1,000 physical exams it was administering previously.
To achieve this, it uses
Amazon Elastic Container Service (Amazon ECS) to orchestrate and automatically scale its Docker containers. The startup has reported no downtime since its migration, during exams or regular operations. Latency has remained around 500 milliseconds during online exams, well below the 1-second limit set internally. “Since migrating to AWS, we’ve been able to deliver a seamless user experience with no interruptions”, adds Mangudkar.
Autoscaling Simplifies Development and Deployment
Previously, when WeShine Tech acquired a new customer or needed to increase user capacity, its developers had to manually provision a new server with a separate domain. “From a DevOps perspective, it was difficult and time consuming. But after we containerized UniApps and started using Amazon ECS, we could better anticipate and prepare for peak loads,” Mangudkar says. Now, it takes just one DevOps team member to launch an initial set of container nodes based on expected exam attendance, and Amazon ECS handles scaling for the remainder of the 12-hour exam window.
Even with a 350 times’ increase in exam loads, WeShine Tech didn’t need additional DevOps engineers due to leveraging
Amazon Elastic Container Registry (Amazon ECR). “We’re not overloaded despite the exponential increase in users. In fact, scaling up for exams is now easier because we’ve integrated Amazon ECR with autoscaling. Our development and deployment cycle, alongside productivity, have improved significantly on AWS,” says Mangudkar.
Multi-Tenant Architecture for Global Expansion
Once UniApps is all-in on the AWS Cloud, WeShine Tech plans to reconfigure its solution for a multi-tenant architecture. It aims to further refine availability and stability as the business looks to rapidly expand abroad. Mangudkar says, “Now that we’ve achieved the scale and correct configurations on the AWS Cloud, we’re focused on adding hundreds, or even thousands, of universities in the next couple of years.”
The company aims to set up joint ventures across the globe, starting in Africa and then the Asia-Pacific region. Mangudkar concludes, “The primary advantage of our solution is the low infrastructure requirement for users, which will appeal to institutions in developing countries.”
Learn More
To learn more, visit aws.amazon.com/machine-learning.
About WeShine Tech
WeShine Tech is a company in India focused on university examination process automation. Its flagship product, UniApps, offers 10 modules and is used by more than 20 domestic universities to manage the end-to-end student lifecycle.
Benefits
- Scales to handle 350,000 concurrent exams, up from 1,000
- Achieves latency of 500 milliseconds during online exams
- Handles large increases in user base without hiring extra staff
- Develops identity verification feature in 1 week
- Authenticates users throughout exams and detects fraudulent activity
- Integrates with external applications such as Zoom
AWS Services Used
Amazon Elastic Container Service
Save up to 50 percent on compute costs with autonomous provisioning, auto-scaling, and pay-as-you-go pricing.
Amazon Rekognition
Amazon Rekognition offers pre-trained and customizable computer vision (CV) capabilities to extract information and insights from your images and videos.
AWS WAF – Web Application Firewall
AWS WAF is a web application firewall that helps protect your web applications or APIs against common web exploits and bots that may affect availability, compromise security, or consume excessive resources.
Amazon Simple Storage Service
Amazon Simple Storage Service (Amazon S3) is an object storage service that offers industry-leading scalability, data availability, security, and performance.
Get Started
Companies of all sizes across all industries are transforming their businesses every day using AWS. Contact our experts and start your own AWS Cloud journey today.