🧠 Shortlistr – AI-Powered Resume Screening

🔥 Inspiration

HRs have to manually go through hundreds of resumes for every job opening. This is time-consuming, repetitive, and prone to bias. I wanted to automate this process and make it smarter and faster using AI.

🚀 What it does

Shortlistr takes a job description and a candidate’s resume, then:

• Analyzes the resume with the job description
  
• Calculates a match percentage
  
• Extracts the candidate’s email
  
• Lets HR send a personalized email to the candidate — all in one click

🛠️ How we built it

• Frontend: Built with React.js and hosted on AWS S3
  
• Backend: Resume analysis logic written in Node.js, deployed on AWS Lambda
  
• Mailing: Email system runs on an EC2 instance using AWS SES
  
• AI: Resume/job match is powered by Gemini API (Google Generative AI)
  
• Used Busboy for parsing file uploads and pdf-parse for extracting text from PDFs

🧱 Challenges we ran into

• Parsing multipart form data in Lambda (no Express!)
  
• pdf-parse failing silently on invalid file buffers
  
• Handling CORS between S3 frontend and Lambda
  
• AWS SES sandbox limitations (can only send to verified emails)

🏆 Accomplishments that we're proud of

• Built and deployed a fully working end-to-end system in under 48 hours
  
• Made everything serverless and production-grade with AWS
  
• Integrated AI to make resume screening intelligent, not just keyword-based

📚 What we learned

• Working with AWS Lambda without Express.js
  
• Handling file uploads and buffers in serverless environments
  
• How to use AWS SES for automated emails
  
• How to design a simple, useful UX that solves a real-world bottleneck for HRs

🔮 What's next for Shortlistr

• Bring the SES setup out of sandbox to enable emailing any candidate
  
• Add bulk resume upload and batch screening
  
• Track email delivery and open status
  
• Provide recommendations to HRs based on candidate fit

Built With

• amazon-ec2
• amazon-web-services
• javascript
• lambda
• node.js
• react
• s3