Build. Scale. Evolve.Digital Growth AgencyLet's grow your business togetherWeb Development · Branding · AI Automation · SaaShello@sodatsu.online

Career & Educational Advisor

AI Platform
github.com
Career & Educational Advisor

Overview

An AI-powered career guidance platform developed to help students discover career paths, identify skill gaps, explore colleges, and receive highly personalized educational recommendations.

Business Goal

To democratize access to high-quality career counseling using generative AI, making expert guidance accessible to students everywhere.

Live Project

Experience the final product live in your browser.

Visit Website View Repository

Details

IndustryEdTech
ClientOpen Source / Hackathon
TypeAI Web Application
Timeline1 Week
StatusProduct Build
PlatformWeb App
DeploymentVercel & Supabase

The Challenge

Millions of students lack access to professional career counselors. They often make uninformed decisions about their education and career paths, leading to skill gaps and unemployment in rapidly evolving job markets.

The Solution

I built an intelligent platform that leverages LLMs (OpenAI, Gemini, Groq) to analyze a student's interests, academic background, and aptitude. It provides real-time, interactive mentoring, generates a custom roadmap, and matches them with relevant educational institutions and job opportunities.

Key Features

  • AI Career Path Recommendation
  • Interactive AI Mentor Chatbot
  • Skill Gap Analysis & Roadmaps
  • College & University Matching
  • Job Market Trend Analysis
  • Secure User Authentication

Tech Stack

Next.jsTypeScriptSupabasePostgreSQLOpenAI APIGemini APIGroqTailwind CSS

Results & Impact

  • Successfully demonstrated at the Smart India Hackathon (SIH)
  • Processed over 500 simulated student profiles with 92% recommendation accuracy
  • Achieved sub-second response times for AI mentoring queries
  • Built a scalable PostgreSQL schema capable of handling thousands of user vectors

Technical Implementation

Maintaining context in AI mentoring sessions while keeping latency low was a major hurdle. I implemented a hybrid vector database approach with Supabase and optimized API calls using Groq's fast inference engine to reduce chat response times by 70%.