Hire me Club — AI-Powered Candidate Screening

Mist Avinya designed and deployed a fully serverless GenAI recruitment platform on AWS, enabling Hire me Club to automate candidate screening, improve recruiter efficiency, and scale hiring operations using Amazon Bedrock Agents and Claude 3.5 Sonnet.

94%Reduction in time-to-shortlist

75%Reduction in recruiter screening time

93%Agent-to-recruiter alignment

About the Customer

Business Context & Challenges

Business Context

Hire me Club is an SMB HR technology company processing 800–1,200 candidate CVs per open role while operating with a lean recruitment team. Manual screening created operational bottlenecks and delayed candidate engagement.

Key Challenges

Recruiter burnout from repetitive manual screening · 5–7 day average time-to-shortlist · inconsistent evaluation standards · difficulty scaling without additional headcount · growing SLA pressure.

Solution

How the Agent Works

Agent Workflow

Retrieves job descriptions and scoring rubrics from DynamoDB · fetches CVs from S3 · extracts structured text with Amazon Textract · evaluates candidates with Claude 3.5 Sonnet · generates ranked shortlists with reasoning · notifies recruiters via SNS.

Key Features

Autonomous multi-step reasoning · human-in-the-loop review for low-confidence candidates · end-to-end trace logging · bias reduction via Bedrock Guardrails · serverless-first architecture · per-role scoring customization.

AWS Services

What Powers It

Bedrock Agents

Core orchestration — reasoning, tool invocation, session management.

Claude 3.5 Sonnet

Candidate evaluation, multi-step reasoning, scoring, justification.

Amazon Textract

Structured text extraction from scanned PDF/DOCX CVs.

Amazon DynamoDB

Job descriptions, rubrics, evaluations, audit trails.

AWS Lambda

Least-privilege serverless functions for every agent tool.

Amazon S3

Secure CV storage with versioning and object locking.

Business Outcomes

Results at a Glance

Time-to-Shortlist

5–7 business days → 4 hours average (94% reduction)

Recruiter Screening Time

65% of working hours → 16% (75% reduction)

Agent-to-Recruiter Alignment

Manual process → 93% alignment (target exceeded)

Peak Hiring Capacity

1x capacity → 3.2x capacity, no added headcount

Architecture

High Availability & Scalability

All services deployed across multiple Availability Zones — Bedrock Agents, Lambda, S3, DynamoDB, SNS, Textract, and API Gateway — with no single point of failure.

Have a similar screening workflow in mind?

Tell us what you're working on. We'll share how we'd approach it.

Start a Conversation

← Back to GenAI & Machine Learning