"AI-Powered Job Search & Career Assistant – Your Gateway to Smarter Job Hunting!"

AI-Powered Job Search & Interview Coach

Inspiration

The job search process can be daunting—tailoring resumes, finding relevant openings, and preparing for interviews is time-consuming and overwhelming. We wanted to streamline and simplify this journey by building an AI-powered platform that assists job seekers at every stage.

Our inspiration came from:

The frustration of repetitive applications and generic cover letters.
The need for real-time feedback during interview practice sessions.
The potential of AI to personalize and enhance the job-seeking experience.

By combining resume parsing, job matching, cover letter generation, and interview coaching, we aimed to create a one-stop solution for job seekers.

What It Does

Our platform offers an end-to-end job search assistant with the following features:

Resume Parsing:

Users upload their resume, and the platform automatically extracts skills, experience, and education using Azure Form Recognizer aka Azure Document Intelligence.
Parsed data is displayed on the dashboard for quick review.

Job Recommendations:

The platform suggests relevant job listings by matching resume details with job titles fetched via a public job API.
Users can bookmark jobs, search, and filter results.

AI-Powered Cover Letter Generator:

Users can generate personalized cover letters with a single click using Llama.
The AI tailors the cover letter to the user's resume and the job description.

Interview Coaching:

Users can practice answering interview questions in real-time.
Speech-to-text transcription captures their responses.
A session summary displays all answers with feedback.

How We Built It

Tech Stack

Frontend: Next.js (TypeScript), TailwindCSS, Shadcn Components
Backend: Drizzle ORM, Neon PostgreSQL
AI Services:
- Azure Document Intelligence → Resume parsing
- Llama using Groq → Cover letter generation
- Azure Speech SDK → Real-time transcription
Authentication: Next-Auth.js with JWT session handling and OAuth.

Development Process

1. Authentication & Dashboard:

Implemented Next-Auth.js for secure authentication with PostgreSQL session handling.
Built a responsive dashboard featuring resume uploads, job widgets, and navigation links.

2. Resume Parsing:

Integrated Azure Document Intelligence for parsing PDF and DOCX resumes.
Extracted and displayed key resume information after formating it through Llama.

3. Job Recommendations:

Fetched job listings of past 7 days from a public job API.
Matched resume skills with job title.
Built a dynamic job listing UI with search, filter, and bookmark options.

4. AI-Powered Cover Letter Generator:

Used Llama with GroqAPI to generate tailored cover letters.
Displayed the cover letter in a preview section with copy, edit and export as pdf options.

5. Interview Coaching:

Implemented speech-to-text transcription using Azure Speech SDK.
Displayed feedback with a session summary.

Challenges We Ran Into

1. Resume Parsing Accuracy:

Inconsistent resume formats made extraction difficult.
We optimized the Document Intelligence configurations to improve parsing accuracy.

2. Real-time Transcription Issues:

Capturing and displaying transcribed responses smoothly was tricky.
We fine-tuned the speech-to-text parameters and added buffering to prevent lag.

3. API Rate Limits:

Public job APIs had rate limits, which restricted the number of searches.

Accomplishments That We're Proud Of

Seamless AI Integration:

Successfully combined Azure Document Intelligencer, Llama, and Speech SDK into a single platform.

Dynamic Job Matching:

Implemented a dynamic job listing system with resume-based recommendations(job title).

Real-time Feedback:

Delivered feedback on interview performance through sentiment analysis through Llama.

User-Centric Design:

Built a clean, responsive, and intuitive UI using TailwindCSS and Shadcn components.

What We Learned

AI Service Optimization:

Fine-tuning AI models is essential for accuracy.
Improved resume parsing reliability by refining Azure configurations.

Efficient State Management:

How to manage states across the application.

Hackathon Execution:

We learned how to prioritize features and manage time effectively under tight deadlines.

What's Next

Enhanced Job Search:

Integrate LinkedIn’s API for real-time job recommendations with proper skill matching.
Add advanced filtering options (e.g., salary range, experience level).

Smart Interview Analysis:

Add emotion and confidence detection during interview sessions.
Provide score-based feedback with improvement suggestions.

Full Deployment & User Testing:

Deploy the platform on Vercel for public access.
Gather user feedback for further improvements.

How I used Github Copilot in VSCode

Accelerating Development and resolving bugs

While coding in Typescript I ran into many type errors and with copilot it was easy to work my way through it by inferring the types automatically using copilot.

Generate Boilerplate code