Automating Recruitment Operations with GenAI on AWS for Elite Group Holding
Automating Recruitment Operations for Elite Group Holding
Use Case - AWS GenAI Implementation
Industry: Financial Services
Geography: India
Employee Size: 100+
Solution: AWS
About Customer
Elite Group Holding is a mid-to-large enterprise with a high demand for continuous hiring across technical and business roles. Managing large volumes of candidate applications while maintaining hiring quality, speed, and consistency had become an increasingly complex challenge for the organization. Talent acquisition, being a business-critical function, directly impacted delivery timelines, project execution, and overall organizational performance.
Challenges
Manual & Inconsistent Job Descriptions
Job descriptions were created manually by HR and hiring managers without standardization, resulting in inconsistent quality and mismatched candidate profiles.
Time-Consuming, Subjective Resume Screening
Hundreds to thousands of resumes were screened manually, making shortlisting time-intensive and prone to human bias, with no intelligent tools available for accurate candidate-job matching.
Disconnected Feedback & Tracking
Interview feedback was collected via spreadsheets and emails, with no centralized system to track candidate status or hiring progress, and limited visibility into recruitment data for decision-making
Scalability & Security Gaps
The absence of a secure, seamless authentication mechanism and real-time monitoring made it difficult to scale recruitment operations securely as hiring demand increased.
Our Approach
enreap designed and implemented a GenAI-powered recruitment solution on AWS for Elite Group Holding, built to automate the hiring lifecycle end-to-end and bring intelligent decision-making into every stage of recruitment. The solution centers on Amazon Bedrock for AI-driven content generation and candidate evaluation, and Amazon Kendra for contextual resume analysis, orchestrated through Amazon EC2 as the central application layer. Supporting services, including Amazon S3, Amazon DynamoDB, Amazon Cognito, and Amazon SNS, ensure secure, scalable, and efficient data handling and communication, while Amazon CloudWatch provides continuous monitoring and observability across the platform. As a result, Elite Group Holding achieved a significant reduction in manual effort and hiring turnaround time, improved candidate quality through AI-driven insights, and established a standardized, scalable, and data-driven recruitment process.
Our Solution
enreap designed and implemented a GenAI-powered recruitment automation platform on AWS for Elite Group Holding, as illustrated in the architecture diagram, enabling end-to-end automation of the hiring lifecycle using scalable, secure, and AI-driven services.
1. User Authentication & Application Access
Recruiters securely access the application, with authentication handled via Amazon Cognito, which issues access tokens for authorized use. User requests are routed through Route 53 → ALB → EC2, with Amazon EC2 acting as the central orchestration layer for API processing, business logic, and integration with AI and data services.
2. Automated Job Description Generation
Recruiters submit job requirements via the UI, and EC2 prepares a structured prompt sent to Amazon Bedrock — using Mistral-Large-3-675B-Instruct as the primary model and NVIDIA-Nemotron-Nano-12B-V2 as the secondary model. Bedrock generates an optimized job description with defined skills and responsibilities, stored in Amazon S3, with a notification sent via Amazon SNS and logs captured in Amazon CloudWatch.
3. AI-Powered Resume Screening & Shortlisting
Resumes uploaded via the application are stored securely in Amazon S3. EC2 sends the resume data to Amazon Kendra, which extracts skills, experience, and contextual insights. EC2 then prepares a composite prompt — combining resume content, the job description, and Kendra insights — and sends it to Amazon Bedrock, using Mistral-Large-3-675B-Instruct for primary reasoning and evaluation and NVIDIA-Nemotron-Nano-12B-V2 for lightweight inference and scoring support. The output includes a candidate summary, skill match score, skill gap analysis, and a shortlist/reject recommendation, with results stored in Amazon DynamoDB (raw documents remain in Amazon S3), the recruiter notified via Amazon SNS, and metrics captured in Amazon CloudWatch.
4. Automated Interview Feedback & Evaluation
An interview transcript is uploaded via the UI along with interviewer ratings. The transcript is stored in Amazon S3, with candidate data retrieved from DynamoDB and the job description retrieved from S3. Amazon Kendra processes the transcript to extract key discussion points, demonstrated skills, and experience insights. EC2 then prepares a structured prompt — including the transcript, candidate data, job description, and interviewer ratings — processed by Amazon Bedrock using Mistral-Large-3-675B-Instruct for deep analysis and feedback generation and NVIDIA-Nemotron-Nano-12B-V2 for fast scoring and classification. The output includes candidate strengths and weaknesses, communication analysis, technical evaluation, a hiring recommendation, and an AI-generated rating score, with final results stored in Amazon DynamoDB, a notification sent via Amazon SNS, and logs captured in Amazon CloudWatch.
5. Secure, Scalable Architecture
AWS IAM ensures role-based access control, and AWS KMS provides encryption for S3, DynamoDB, and SNS. A private subnet architecture ensures secure processing, with a NAT Gateway enabling controlled outbound communication.
Business Outcome
1. Reduced resume screening effort by 70–80% through AI-powered automation
2. Achieved a 40–60% faster hiring cycle, significantly improving time-to-hire
3. Improved quality of hire through standardized, data-driven candidate evaluation
4. Eliminated manual effort in job description creation, screening, and feedback generation
5. Enabled unbiased, consistent candidate assessment using AI-driven scoring models
6. Enhanced recruiter productivity by automating repetitive tasks
7. Improved candidate experience through faster, automated communication
8. Centralized recruitment data with real-time visibility and monitoring
9. Scaled seamlessly to handle high volumes of candidate applications
10. Improved decision-making speed with AI-generated insights and recommendations
11. Reduced operational costs associated with manual recruitment processes
12. Reduced operational costs associated with manual recruitment processes
13. Established a secure, compliant recruitment system aligned with AWS best practices