The Challenge

DBS hires across a broad range of customer-facing banking functions where success depends on more than technical qualifications and previous experience.

Candidates may have similar credentials and professional backgrounds while differing significantly in communication ability, customer orientation, stakeholder management, problem-solving and relationship-building capabilities.

Traditional recruitment processes relied heavily on resumes and interviews, making it difficult to consistently identify these qualities early in the hiring process. Recruiters and hiring managers needed deeper candidate intelligence to make better shortlisting decisions without adding more manual evaluation effort.

The challenge was to improve the quality and speed of talent evaluation while maintaining human oversight throughout the hiring process.

The Solution

DBS leveraged Savos by impress.ai, an AI-powered hiring platform, to add an intelligent candidate investigation layer to its recruitment process.

Rather than relying solely on information contained in a resume, Savos uses AI-led candidate conversations to investigate relevant capabilities and gather additional evidence about a candidate’s experience and suitability.

ScaleScreen, the AI screening capability within Savos, engages candidates through structured and adaptive conversations designed around the requirements of the hiring process. It can explore areas such as communication, stakeholder management, customer engagement, problem-solving and relevant professional experience.

The resulting information is then synthesized through TalentLens into structured candidate intelligence that recruiters and hiring managers can use during shortlisting and interview preparation.

This approach allowed DBS to move from evaluating candidates primarily on what was already documented on their resumes to making decisions using a richer view of their capabilities and experience.

How AWS Powered the Solution

Savos was deployed using AWS infrastructure to provide the scalable compute, AI, data, messaging, networking, security and monitoring capabilities required to operate the solution.

Amazon Bedrock provides the foundation for the generative AI capabilities used within Savos to support candidate investigation and generate candidate intelligence.

The application layer uses Amazon ECS and Amazon EC2 for compute workloads, while Amazon RDS and Amazon ElastiCache support persistent data and caching requirements. Amazon SQS supports asynchronous communication between application components.

The solution uses Elastic Load Balancing and Auto Scaling to distribute workloads and scale application capacity based on demand. Amazon CloudFront, Amazon Route 53 and Amazon S3 support application delivery and related infrastructure requirements.

Security and operational controls include AWS KMS, VPC endpoints, security groups, NAT Gateway, AWS WAF and Amazon CloudWatch. The architecture uses Multi-AZ deployment, including Multi-AZ configurations for relevant data services, to support resilience and availability.

This AWS architecture enables Savos to support AI-powered candidate evaluation workloads while providing the scalability, reliability, security and operational controls required for enterprise recruitment environments.

The Impact

The implementation delivered measurable improvements across the recruitment evaluation process:

  • 55% reduction in recruiter screening effort
  • 3x faster candidate evaluation workflows
  • 35% improvement in shortlisting consistency across hiring teams
  • 25% increase in recruiter productivity
  • 40% faster time-to-shortlist

The solution also provided recruiters and hiring managers with richer candidate intelligence earlier in the recruitment process, helping them spend more time on high-value hiring decisions and less time on manual candidate investigation.