Build an AI Jobs Agent with Amazon Bedrock AgentCore | End-to-End AWS Project | Episode26

Build a complete AI Jobs Agent using Amazon Bedrock AgentCore, Amazon Nova, and AWS serverless services. In this end-to-end AWS project, I demonstrate how to build, secure, deploy, and operate an agentic application that analyzes job descriptions, evaluates résumé evidence, generates an ATS-style score, creates a tailored résumé and cover letter, performs an independent review, and recommends the candidate’s next best action. 🚀 Try the live AWS Jobs Agent: https://jobs.knowledgebytes.info/ 🌐 Visit Knowledge Bytes: https://knowledgebytes.info/ 💻 Source code: https://github.com/mahendra15nov/auto... WHAT YOU WILL LEARN ✅ What Amazon Bedrock AgentCore is ✅ Chatbot versus agentic application ✅ How AgentCore Runtime executes an AI agent ✅ How to design a multi-step agentic workflow ✅ Résumé and job-description extraction ✅ ATS-style evidence matching ✅ Tailored résumé and cover-letter generation ✅ Independent AI review and next-action recommendations ✅ Human-in-the-loop application workflow ✅ User registration and authentication with Amazon Cognito ✅ Securing APIs with API Gateway and JWT authorization ✅ Serverless backend development with AWS Lambda ✅ Private per-user storage with Amazon DynamoDB ✅ Secure résumé uploads using Amazon S3 presigned URLs ✅ Hosting a private S3 frontend through Amazon CloudFront ✅ Connecting jobs.knowledgebytes.info with Route 53 and ACM ✅ Monitoring Lambda and agent errors through CloudWatch ✅ Controlling token consumption and public AI usage ✅ Deploying AWS infrastructure with SAM and CloudFormation CHAPTERS 00:00 Introduction 01:10 What We Will Build 02:25 AWS Jobs Agent Live Demo 07:30 Product and User Journey 09:20 Chatbot vs Agentic Application 11:15 Amazon Bedrock AgentCore Explained 14:00 AgentCore Capabilities 16:20 Complete AWS Architecture 19:10 Agentic Analysis Workflow 21:40 Prompt Security and Responsible AI 24:00 User Data Isolation 25:50 Token and Cost Protection 28:10 Source Code Walkthrough 34:30 AWS Console Walkthrough 40:30 Deployment Pipeline 42:15 Roadmap and Closing AWS ARCHITECTURE The application uses: • Amazon Bedrock AgentCore Runtime • Amazon Nova • Amazon Cognito • Amazon API Gateway • AWS Lambda • Amazon DynamoDB • Amazon S3 • Amazon CloudFront • Amazon Route 53 • AWS Certificate Manager • AWS CloudFormation and SAM • Amazon CloudWatch APPLICATION WORKFLOW 1. The user creates a secure Cognito account. 2. The user uploads a PDF, DOCX, or TXT résumé. 3. The application extracts the candidate’s profile and skills. 4. The user imports or pastes a target job description. 5. AgentCore compares job requirements with résumé evidence. 6. The agent generates an ATS-style score, strengths, and gaps. 7. It prepares a tailored résumé and cover letter. 8. An independent reviewer evaluates the application package. 9. The agent recommends the next best action. 10. The user reviews, downloads, submits, and tracks the application. This is not an automatic job-application bot. The agent prepares and recommends, while the user reviews and approves every important action. SECURITY AND RESPONSIBLE-AI FEATURES • Private, user-isolated application records • Cognito JWT authentication • Server-side input validation • Evidence-only résumé generation • Protection against invented qualifications • Private résumé storage • Short-lived S3 upload permissions • Duplicate-analysis reuse • Configurable usage and token controls • Human approval before applying If you found this project useful, please LIKE the video, SUBSCRIBE to Knowledge Bytes, and share your questions in the comments. Keep learning something new—one byte at a time. #AmazonBedrock #AgentCore #AWS