Background

With AI-assisted shortlisting and InterviewMate, SIT streamlined candidate screening, improved consistency in evaluation, and reduced administrative workload. The platform has played an important role in supporting our digital transformation efforts in their admissions process.

To support this transformation, Singapore Institute of Technology (SIT) partnered with impress.ai, an AWS Partner, to deploy an AI-powered admissions platform built on Amazon Web Services (AWS). The solution combines scalable cloud infrastructure with AI-powered admissions capabilities to help admissions teams evaluate applicants more efficiently while maintaining a secure, resilient, and highly available platform during peak admission periods.

The Problem

SIT required a scalable admissions screening platform to modernize admissions workflows, improve candidate engagement, and reduce dependency on manual screening and coordination processes.

Traditional admissions operations created:

  • fragmented candidate experiences
  • recruiter coordination overhead
  • inconsistent screening workflows
  • operational bottlenecks during admission cycles

The institution required a centralized AI-powered platform capable of supporting scalable operations.

AWS-Powered Admissions Platform

Built on Amazon Web Services (AWS), the impress.ai platform leverages Amazon Bedrock to power AI-assisted applicant engagement, intelligent shortlisting, and recruiter decision support. Through InterviewMate and AI-powered admissions capabilities, the platform helps admissions officers and faculty evaluate applicants more consistently while maintaining human oversight throughout the admissions process.

The solution runs on a highly available AWS architecture using Amazon ECS, Amazon EC2, Amazon RDS, Amazon ElastiCache, Amazon S3, Amazon CloudFront, and Amazon Route 53, with Amazon SQS orchestrating admissions workflows across the platform. Enterprise-grade security is supported through AWS Key Management Service (AWS KMS), private VPC networking, and Multi-AZ deployment, enabling SIT to securely scale admissions operations while protecting applicant data and maintaining high availability during peak application cycles.

Outcomes of Project & Success Metrics

Initial deployment indicators included:

  • 40–60% reduction in recruiter coordination effort
  • 2x faster candidate screening workflows
  • improved candidate engagement across admission stages
  • improved recruiter productivity
  • reduced workflow bottlenecks during admissions cycles

Built on AWS, the platform enables SIT to securely support large-scale admissions while delivering faster applicant evaluations, reducing administrative effort, and providing a more consistent admissions experience for applicants, admissions officers, and faculty members.

KPIs being tracked include:

  • application processing time
  • recruiter workload reduction
  • workflow completion rates
  • candidate engagement metrics
  • candidate drop-off reduction

Lessons Learned

Key lessons learned included:

  • conversational engagement improved candidate participation rates
  • workflow orchestration improved recruitment visibility
  • centralized screening improved evaluation consistency
  • scalable infrastructure was critical during peak admission periods