PlacementPilot AI

PlacementPilot AI: An autonomous career agent that analyzes job opportunities, identifies skill gaps, creates study plans, and guides students from preparation to placement.

Comment

🚀 Inspiration

Placement preparation is one of the most stressful and fragmented experiences for students. Every application requires analyzing job descriptions, identifying missing skills, creating study plans, tracking progress, and staying consistent throughout the journey.

We noticed that students spend more time organizing their preparation than actually learning. Existing AI tools mostly act as chatbots that answer questions, but they don't take ownership of the workflow itself.

We wanted to build an AI system that does the heavy lifting automatically. Our vision was simple: upload a resume, choose a target role, and let autonomous AI agents handle the planning, analysis, and monitoring in the background.

This idea became PlacementPilot AI.

🎯 What it does

PlacementPilot AI is an autonomous, multi-agent career preparation platform powered by Gemini and Google ADK.

Users simply upload their resume and choose a target role. The system then launches an autonomous workflow that:

  • Analyzes the resume
  • Extracts existing skills
  • Analyzes target job requirements
  • Identifies skill gaps
  • Generates a personalized 30-day preparation roadmap
  • Creates daily learning tasks
  • Tracks progress over time
  • Monitors completion rates
  • Generates reminders and recommendations automatically

Unlike traditional AI assistants, PlacementPilot AI does not rely on continuous user prompts. The platform operates through asynchronous background workflows that continue working even after the user leaves the application.

🏗️ How we built it

We designed PlacementPilot AI as a cloud-native, event-driven multi-agent system.

Frontend

  • Next.js 15
  • TypeScript
  • Tailwind CSS
  • ShadCN UI

Backend

  • FastAPI
  • Python

AI Layer

  • Gemini 3.5
  • Google Agent Development Kit (ADK)

Google Cloud Services

  • Cloud Run
  • Firestore
  • Pub/Sub
  • Cloud Scheduler
  • Firebase Authentication

The system consists of four specialized AI agents:

Resume Analysis Agent

Extracts technical skills, strengths, weak areas, and ATS-related insights from uploaded resumes.

Job Intelligence Agent

Parses target job descriptions and identifies required and preferred skills.

Study Planner Agent

Builds a personalized 30-day preparation roadmap and generates daily learning tasks.

Progress Monitoring Agent

Runs automatically every day, tracks completion status, logs progress, and generates reminders.

The workflow is orchestrated through Pub/Sub events and runs asynchronously in the background, while Firestore stores long-term state and progress data. The entire platform is deployed on Google Cloud Run for scalability and cost efficiency.

⚠️ Challenges we ran into

One of the biggest challenges was moving beyond a traditional chatbot architecture.

Most AI projects stop at generating responses, but we wanted our system to execute meaningful workflows autonomously. Designing a reliable multi-agent pipeline required careful separation of responsibilities between agents while maintaining shared context and state.

Another challenge was implementing asynchronous execution. We needed the frontend to remain responsive while the AI agents performed computationally intensive tasks in the background. Pub/Sub orchestration and Firestore state management helped us solve this challenge.

Ensuring consistency across multiple agents and preventing workflow failures also required robust orchestration and structured outputs from Gemini-powered agents.

🏆 Accomplishments that we're proud of

  • Built a true autonomous AI workflow rather than a chatbot.
  • Successfully implemented a multi-agent architecture using Google ADK.
  • Created an event-driven system powered by Pub/Sub.
  • Integrated Gemini 3.5 into a real-world productivity use case.
  • Designed a scalable cloud-native architecture on Google Cloud.
  • Implemented long-term memory and state persistence using Firestore.
  • Developed a solution that addresses a real challenge faced by millions of students preparing for placements.

Most importantly, we transformed placement preparation from a manual process into an autonomous workflow.

📚 What we learned

Through this project, we learned that effective AI systems are not just about generating intelligent responses—they are about taking action.

We gained hands-on experience with:

  • Multi-agent architecture design
  • Google Agent Development Kit (ADK)
  • Gemini 3.5 integration
  • Event-driven cloud systems
  • Firestore data modeling
  • Cloud Run deployment
  • Asynchronous workflow orchestration
  • State management for long-running AI systems

We also learned how to design AI systems that remain maintainable, scalable, and production-ready.

🔮 What's next for PlacementPilot AI

Our vision is to evolve PlacementPilot AI into a complete AI-powered career operating system.

Future enhancements include:

  • LinkedIn integration
  • Automatic job discovery
  • AI-powered mock interviews
  • Coding assessment generation
  • Personalized resume optimization
  • Industry-specific preparation tracks
  • Recruiter insights and analytics
  • Career mentor collaboration features
  • Interview scheduling assistance
  • Real-time placement readiness scoring

We believe the future of career preparation is autonomous, personalized, and continuously adaptive. PlacementPilot AI is our first step toward making that future a reality.

Built With

  • agentic-ai
  • artificial-intelligence
  • career-tech
  • cloud-scheduler
  • edtech
  • fastapi
  • firebase-auth
  • firestore
  • gemini
  • gemini-3-5
  • generative-ai
  • google-adk
  • google-cloud
  • google-cloud-run
  • google-pubsub
  • multi-agent-systems
  • nextjs
  • nextjs-15
  • python
  • react
  • shadcn-ui
  • tailwindcss
  • typescript
  • vertex-ai
Share this project:

Updates

Submission history