CareerTwin AI

• 

  areerTwin AI Interactive Dashboard — Instantly simulating alternate 10-year parallel career tracks from a single resume upload.

Inspiration

Every professional faces a "What If?" moment in their life. What if I chose a hyper-growth startup over a comfortable FAANG corporate role? What if I abandoned the standard track to become a digital nomad? CareerTwin AI was inspired by this exact human curiosity. We wanted to build an engaging, data-driven engine that acts as a mirror to your potential, letting you safely simulate and navigate 4 completely distinct 10-year alternate professional futures.

How We Built It

The application is built on a modern, high-performance full-stack architecture:

• Frontend: Built using Next.js (App Router) and TypeScript to create a lightning-fast Single Page Application dashboard interface. We leveraged Tailwind CSS to design a dark, glassmorphism-themed interactive cyberpunk aesthetic.
• Backend: Powered by Node.js and Express.js to handle asynchronous API calls, heavy simulation calculations, and state tracking.
• Database & ORM: Implemented Prisma ORM coupled with a relational layout to handle structural profiles, metric logs, and generated timelines smoothly.

To model user scoring profiles dynamically across distinct timelines, we used a customized career compound scoring formula: $$S_{\text{trajectory}} = \sum_{t=1}^{10} \frac{E_t \cdot T_t \cdot L_t}{(1 + \gamma)^t}$$

Where $E_t$ represents Employability, $T_t$ represents Technical Depth, $L_t$ represents Leadership capabilities at year $t$, and $\gamma$ represents the risk-factor attenuation associated with sudden career domain pivots.

Challenges We Faced

One of our biggest hurdles was architectural synchronization. Managing a full-stack monorepo (frontend and backend inside a single repository) introduced deployment challenges on cloud container platforms, especially when coordinating environment variables and root path configurations.

Additionally, optimizing the file stream pipelines to parse physical document data via memory storage arrays while instantly calculating timeline metrics required deep iterative testing to eliminate latency bottlenecks.

What We Learned

Through building CareerTwin AI, we mastered structural multi-service application design and production environment optimization. We learned how to write clean, maintainable relational data schemas using Prisma and how to effectively map tabular timelines into complex visual node paths in React. Most importantly, we learned how to present sophisticated AI data in an interface that is clear, accessible, and deeply engaging for everyday users.

Built With

• express.js
• git
• multer
• next.js
• node.js
• postgresql
• prisma
• sqlite
• tailwind-css
• typescript