scroll down ↓
Will AI Take My Job? [2026]
A retro web app I built independently with AI-assisted development, making AI displacement risk searchable across 1,011 occupations.
psst, the stickers can be moved: drag them around












An independent digital product built around a simple idea: make the changing landscape of AI and employment easier to explore. The application transforms a government occupational dataset into an interactive Windows 98-style file explorer, organising 1,011 occupations across 32 career clusters.
The interface takes inspiration from an operating system most people recognise. Instead of presenting AI displacement as another alarming infographic or corporate prediction, it gives users a familiar environment to explore the information, look up their own occupation and discover other career paths.
Users can browse occupations by cluster, search job titles using fuzzy matching (including common synonyms), and filter results by all jobs, at-risk roles or those classified as safe. Each occupation includes its official description alongside its AI-displacement status. 90 roles are flagged as at risk.
I developed the application independently using AI-assisted development, working with React, Vite and Tailwind for the interface, Fuse.js for search, and Python and openpyxl for data processing. The application is deployed on Vercel and draws on official US occupational classifications (SOC codes).
The shipped version established the foundation for a broader career-guidance platform I was developing, with planned features including AI-powered risk explanations, skills-gap assessments, career-pivot suggestions and university recommendations.
What it taught me: how to take an existing dataset and turn it into a usable product, work across the boundaries of design and development, and make deliberate interface decisions around how people experience unfamiliar or potentially intimidating information.

