MatchMeMaybe

Find Your Dream Job with a Swipe

Comment

Inspiration

MatchMeMaybe (🤖 AI Assistants & Automation) was inspired by the success of dating apps and their intuitive swipe-based interface. We recognized that job searching could be made more engaging and efficient by applying similar principles. The traditional job search process often feels overwhelming and time-consuming, so we set out to create a more modern, user-friendly approach that leverages AI and machine learning to match candidates with their ideal opportunities.

What it does

MatchMeMaybe is an AI-powered job matching platform that combines the familiarity of dating apps with sophisticated job matching technology. Key features include:

  • Smart Job Matching: Uses LLMs through AWS Bedrock to analyze resumes and match candidates with relevant job opportunities
  • Swipe Interface: Intuitive swipe-based interface for job discovery (right to save, left to skip)
  • Resume Analysis: Automated resume parsing and analysis using AWS Textract
  • Job Tracking: Comprehensive job application tracking system with status updates using AWS S3 and DynamoDB
  • Interactive UI: Modern, responsive design with animations and visual feedback
  • Status Management: Track jobs through different stages (saved, applied, interviewing, offered)

How we built it

Frontend:

  • HTML and React with TypeScript
  • Framer Motion for animations
  • Tailwind CSS for styling
  • Redux for state management

Backend:

  • AWS Lambda for serverless functions
  • Amazon S3 for file storage
  • Amazon Textract for resume parsing
  • Amazon Bedrock for AI-powered resume parsing and job matching
  • Amazon OpenSearch for job search and indexing
  • Amazon EC2 for retrieving jobs from LinkedIn
  • Amazon IAM to control all permissions

Key AWS Services:

  1. Amazon S3

    • Stores user resumes and profile data
    • Implements CORS for secure file uploads
    • Manages file access and permissions

  2. AWS Lambda

    • Processes resume uploads
    • Handles job matching logic
    • Manages application status updates

  3. Amazon Textract

    • Extracts text from uploaded resumes
    • Parses structured data from documents
    • Identifies key information like skills and experience

  4. Amazon Bedrock

    • Used to invoke prompts on Anthropic's Claude 3.5 model
    • Analyzes job requirements against candidate profiles
    • Extracts key information from the user's resume

  5. Amazon OpenSearch

    • Indexes job listings for efficient and relevant search results
    • Enables advanced job filtering and sorting

  6. Amazon EC2

    • Searches for job listings on LinkedIn as per LLM generated queries
    • Runs based on Lambda request

Challenges we ran into

  1. AWS OpenSearch: Connecting all the AWS services together was a complex task. One of the main issues we had was trying to set up OpenSearch. While we were able to query different indexed jobs in the vector space in the end, we did not have enough time to fully implement it with the website. Instead we opted for a different approach, for each job asking the LLM whether to keep it or not.
  2. AWS Lambda: Programming Lambda and connected it to its services was pretty difficult. Mainly because of permission errors and having to submit the code through a zip file took time to debug.
  3. Full Stack Management: Connecting the very dynamic frontend to the backend as well as connecting it to the cloud was difficult
  4. Hosting the website: While we didn't end up hosting it in the cloud, we made a strong attempt which was very challenging.

Accomplishments that we're proud of

  1. Innovative Interface: Implemented a swipe-based job discovery system with fun features
  2. AI Integration: Built a sophisticated job matching system using AWS Bedrock and Lambda
  3. Real-time Updates: Implemented real-time status updates with visual feedback using DynamoDB

What we learned

  1. AWS Services: Deep understanding of AWS serverless architecture and services
  2. AI/ML Integration: Experience with implementing AI-powered features
  3. State Management: Best practices for managing complex application state
  4. UI/UX Design: Creating engaging user interfaces with modern web technologies
  5. Performance Optimization: Techniques for optimizing real-time applications

What's next for MatchMeMaybe

  1. Company Profiles: Add detailed company profiles and reviews
  2. Mobile App: Develop native mobile applications for iOS and Android
  3. Analytics Dashboard: Add analytics for job seekers to track their application success
  4. Social Features: Implement networking features for job seekers
  5. Account Creation: Add functionality for user accounts to be created and saved
Share this project:

Updates