in progress

StepIn

An AI-powered career exploration platform that helps students discover potential career paths through immersive experiences built from real professional journeys. The platform uses AI agents to transform professional reflections into interactive career worlds and generate personalized Career DNA insights for students.

Next.jsTypeScriptFastAPILangGraphGeminiSupabaseRedisTailwind CSS

Project Overview

StepIn is an AI-powered career exploration platform designed to help students discover potential career paths through immersive, interactive experiences built from real professional journeys.

Instead of relying on generic career quizzes or static job descriptions, the platform uses AI agents to transform professional reflections into interactive career worlds that students can explore.

The system generates personalized Career DNA insights based on a student's interests, strengths, and goals—helping them understand not just what careers exist, but how those careers actually feel day-to-day.

The Problem

Career exploration for students is often overwhelming and impersonal. Most students rely on generic descriptions, outdated information, or random suggestions that don't reflect their actual interests and strengths.

Existing career tools typically offer:

  • Static job listings with minimal context
  • Generic personality quizzes that lack depth
  • Surface-level advice without real-world insight

The challenge was creating a platform that makes career exploration immersive, personalized, and grounded in real professional experiences rather than abstract categories.

Architecture & Tech Stack

The system follows an AI-agent-driven architecture designed around personalized content generation and interactive exploration.

Frontend

Next.js + TypeScript — Building a fast, type-safe application interface with modern web architecture.

Tailwind CSS — Efficient UI development and consistent design systems.

Backend

FastAPI — Provides the API layer and handles communication between the frontend, AI systems, and databases.

AI Orchestration

LangGraph — Used to coordinate multiple AI agents with different responsibilities: career world generation, reflection processing, and Career DNA analysis.

Gemini — Provides the language model capabilities for generating interactive career experiences and personalized insights.

Data & Infrastructure

Supabase — Handles application data, user management, and real-time features.

Redis — Supports fast temporary state management and caching.

Key Features

AI Career World Generation

Transforms professional reflections and real career journeys into interactive worlds that students can explore—making career discovery immersive rather than abstract.

Career DNA Insights

Generates personalized insights based on a student's interests, strengths, and goals—helping them understand how their profile aligns with different career paths.

Multi-Agent AI Workflow

Different AI agents handle different parts of the process: reflecting on professional journeys, generating interactive content, and analyzing student profiles.

Student-Centered Exploration

Designed around how students actually think about careers—through experiences, stories, and personal connection rather than job titles and salary ranges.

Engineering Challenges

Generating Meaningful Career Experiences

The challenge was going beyond simple text generation to create interactive, immersive career worlds that feel authentic and helpful—not just another chatbot response.

Personalization at Scale

Each student has different interests, backgrounds, and goals. The system needed to generate personalized experiences without becoming overwhelming to build or maintain.

Coordinating Multiple AI Agents

The multi-agent workflow required careful orchestration to ensure different agents maintained context and produced coherent, connected experiences.

Lessons Learned

Building StepIn reinforced how powerful AI agents can be when applied to knowledge-intensive workflows. The key insight was that the right architecture—multiple specialized agents working together—produces far more useful results than a single large prompt.

The project also highlighted the importance of grounding AI-generated content in real data. Career exploration is only valuable when it reflects actual professional experiences, not generic descriptions.

Future Work

Future development focuses on expanding the platform's reach and depth:

  • Expanding the library of real professional journeys
  • Improving Career DNA analysis with more nuanced profiling
  • Adding collaborative exploration features for students
  • Building integrations with educational institutions
  • Exploring AR/VR experiences for deeper immersion