Overview
I integrated Amazon Bedrock into a cloud-based recruiting application to deliver intelligent candidate recommendations. Nest.js runs an async prompt pipeline against Claude on Bedrock, with candidate records in PostgreSQL and CVs in S3, while React surfaces ranked recommendations to recruiters.
Architecture
flowchart TB Recruiter(["Recruiter"]) FE["React frontend"] API["Nest.js API"] Queue["Async job / prompt pipeline"] Bedrock["Amazon Bedrock Claude"] PG[(PostgreSQL candidates)] S3[(S3 CVs)] Recruiter -->|"Role requirements + search"| FE FE --> API API --> PG API --> S3 API -->|"Enqueue match job"| Queue Queue -->|"Prompt pipeline"| Bedrock Bedrock -->|"Match results"| Queue Queue --> API API -->|"Ranked recommendations"| FE
Summary
An AI-powered recruiting assistant that recommends candidates by best fit, using Amazon Bedrock and Claude foundation models for semantic matching.
What I worked on
- Integrated Amazon Bedrock into the recruiting application.
- Designed async prompt workflows for semantic candidate–role matching.
- Leveraged Claude foundation models to deliver intelligent search and recommendations over Postgres candidates and S3 CVs.
Results
- AI-assisted candidate recommendations based on best fit
- Semantic matching workflows powered by Claude on Bedrock
- Faster, more informed hiring decisions for recruiters