GetHired

Take back your time and and GetHired with AI

Comment

💡 Inspiration

Job searching is time-consuming, repetitive, and often emotionally exhausting. Between scanning listings, tailoring resumes, researching companies, and tracking applications, candidates spend hours managing the process instead of focusing on interviews or upskilling.
GetHired was built to fix that—by automating the job hunt through a system of autonomous multi-agents that work together to streamline and personalize every step of the process.

🚀 What it does

GetHired is a cloud-based, AI-powered assistant that automates the job application workflow using a system of four specialized agents:

🎯 MVP Features

🕵️‍♀️ Job Discovery Agent

  • Scans job boards (LinkedIn, Indeed, etc.) using keywords, location, and remote filters
  • Prioritizes jobs based on salary, role fit, and company reputation

📝 Resume Tailoring Agent

  • Analyzes job descriptions
  • Suggests resume tweaks to match required skills and keywords
  • Generates personalized resumes for each job
  • Generates personalized cover letters for each job

🧠 Company Research Agent

  • Gathers insights on company culture, Glassdoor reviews, and team structure
  • Summarizes LinkedIn profiles of hiring managers
    • Uses the Google Maps API to present location insights for job seekers looking to relocate

📬 Job Coach

  • Learns user preferences over time
  • Sends reminders for follow-ups and interview prep
  • Logs progress in Firestore

⚙️ How We Built It

Tech Stack:

  • Agent Framework: Custom ADK (Agent Development Kit) in Python
  • Cloud Infrastructure: Google Cloud Run, Firestore, Vertex AI
  • Storage & Messaging: Firestore (job data, resume versions), Agent-to-Agent (A2A) protocol

Architecture:

  • Agents store results in Firestore and publish the next task event
  • Agents run off a standalone FastAPI server for A2A compatibility
  • Followed Google’s Agent-to-Agent (A2A) protocol

🧠 Challenges We Ran Into

  • Deploying agents via Cloud Run due to MCP server timeouts and port conflicts
  • Parsing agent responses in our frontend React application
  • Personalizing resumes accurately across different job types using prompt engineering

✅ Accomplishments We’re Proud Of

  • Learned how to design and deploy a multi-agent system within one month without prior experience
  • Integrated Google Cloud tools for reliable async messaging and data persistence
  • Created a clean agent lifecycle and ADK abstraction
  • Completed an end-to-end demo flow from job discovery to resume tailoring

📚 What we learned

  • How to build modular agents that communicate effectively using Google’s ADK
  • Managing state and context across distributed, asynchronous tasks
  • Integrating our agents into a React frontend
  • Best practices for building serverless apps with Google Cloud
  • How to Build our own MCP sever
  • Sharpened prompt engineering skills for task-specific AI outputs (e.g., resume personalization)

📈 What’s Next

  • A networking agent that gathers connections, reaches out to them automatically upon approval, and keeps track of responses and reminders
  • Google Chrome extension that uses a user’s profile information to make applications less time-consuming

Built With

Share this project:

Updates