AI Career Coach

Inspiration

Job seekers often struggle with three major problems: understanding technical concepts deeply, preparing for interviews with realistic practice, and creating a focused roadmap to reach their target role. Many candidates jump between YouTube videos, blogs, and random interview question lists, which makes preparation fragmented and stressful.

We built AI Career Coach to solve this problem in one place. Our goal was to create an intelligent assistant that helps users learn topics, practice mock interviews, get AI-based answer feedback, and receive a personalized 30-day career plan. We wanted to make interview preparation more interactive, practical, and accessible using generative AI.

What it does

AI Career Coach is an AI-powered web application that helps candidates prepare for technical interviews and career growth.

Key features:

Explain Topic Users can enter any technical topic (for example: Spring Boot, Microservices, Java Collections), and the AI generates a beginner-friendly yet practical explanation.

Mock Interview Question Generator Users provide:

target role

experience level

tech stack The app generates role-specific mock interview questions.

AI Interview Answer Evaluator Users can type or speak their answers, and the AI evaluates them with:

score (1–10)

strengths

areas for improvement

an ideal answer suggestion

Voice Input for Answers Users can answer interview questions using speech recognition, making the experience feel more like a real interview.

AI Interview Scorecard After completing the interview, the app generates a summary scorecard based on all answers provided.

30-Day Career Plan Generator Users enter their target role, company, experience, and tech stack, and the AI creates a structured 30-day preparation roadmap.

Overall, the platform acts like a personalized AI interview coach + learning assistant + career planner.

How we built it

We built AI Career Coach as a full-stack web application with a clean and simple interface.

Tech stack used:

Backend: Java + Spring Boot

Frontend: HTML, CSS, Bootstrap, JavaScript

AI Model: Amazon Nova (via Amazon Bedrock)

Rendering: Markdown parsing for structured AI responses

Voice Input: Web Speech API (SpeechRecognition / webkitSpeechRecognition)

Architecture overview:

The frontend provides separate modules for:

topic explanation

interview question generation

answer evaluation

scorecard generation

career plan generation

The Spring Boot backend exposes REST APIs such as:

/ai/explain

/ai/interview

/ai/evaluate-answer

/ai/interview-summary

/ai/career-plan

These APIs interact with Amazon Nova to generate dynamic AI responses based on user input.

Key implementation details:

Built prompt-driven AI workflows for each feature

Used structured output prompts for interview evaluation and career plans

Added frontend validation to improve user experience

Integrated voice-to-text for mock interview simulation

Implemented question navigation (next / previous) and answer persistence across questions

Challenges we ran into

We faced several practical challenges while building the project:

Prompt consistency with AI output Generative AI does not always return responses in the exact structure expected. We had to refine prompts multiple times to ensure consistent formatting for evaluation, scorecards, and career plans.

Speech recognition accuracy for technical terms Terms like Spring Boot, Microservices, and Java Developer were sometimes misheard or duplicated by browser speech recognition. We had to tune recognition settings and clean the output for a smoother demo experience.

Handling malformed or inconsistent responses Sometimes AI responses included extra text, code blocks, or formatting issues, so we added cleanup and parsing strategies.

Balancing strictness in evaluation If the evaluator was too lenient, weak answers got high scores. If too strict, even decent answers looked poor. We had to tune prompts to make feedback more realistic and useful.

Frontend state management for interview flow Managing question navigation, storing answers per question, and generating the final scorecard required careful handling in JavaScript.

Accomplishments that we're proud of

We’re proud that we built a complete end-to-end AI-powered interview preparation platform instead of just a single chatbot feature.

Highlights:

Successfully integrated Amazon Nova into a real-world use case

Built multiple AI workflows inside one product

Created a realistic mock interview experience with voice input

Designed a personalized answer evaluator with scoring and feedback

Added a 30-day AI-generated career roadmap, which makes the product useful beyond interviews

Built the project as a working full-stack application, not just a prototype concept

What makes us most proud is that the project feels practical and immediately useful for students, freshers, and working professionals preparing for jobs.

What we learned

This project taught us a lot about both engineering and product design:

How to integrate Amazon Bedrock / Amazon Nova into a Spring Boot application

How important prompt engineering is for getting reliable AI output

How to design AI features that are actually useful in a user workflow, not just “AI for the sake of AI”

How to handle response formatting, parsing, and validation when building production-like AI applications

How to combine frontend interactivity with backend AI services

How to create a better user experience by blending:

structured prompts

voice input

guided workflows

actionable AI feedback

What's next for AI Career Coach

We see a lot of future potential for AI Career Coach.

Planned next steps:

Resume Analyzer Upload resume and get ATS-style feedback, keyword suggestions, and role-fit analysis

Company-Specific Interview Mode Generate interview questions tailored for companies like TCS, Infosys, Amazon, Accenture, etc.

Behavioral + HR Interview Practice Add STAR-method feedback for HR and leadership questions

Speech Confidence Analysis Analyze speaking pace, filler words, and confidence for voice answers

User Progress Tracking Dashboard Track scores over time and identify weak areas

Role-based learning paths More detailed plans for Java Developer, Full Stack Developer, Data Analyst, Cloud Engineer, etc.

Deployment as a hosted platform Move from local demo to cloud-hosted version with authentication and saved user sessions

Our long-term vision is to make AI Career Coach a complete AI-powered career preparation companion for learning, practice, and growth.

Built With

• html
• microservice
• springboot