Developer Daniel Balcarek built a space-station themed portfolio for a Google AI challenge, using a terminal interface and AI chatbot. The project, hosted on Google Cloud Run, is fully data-driven, replacing static content with a configurable system. It merges his interests in gaming, sci-fi, and software craftsmanship into an interactive experience.

The technical stack combines an Angular frontend with a .NET backend proxy. This proxy manages prompts and communicates with Gemini models, enabling features like a Job Fit Analyzer. This tool parses job descriptions using an LLM, then scores matches against the portfolio with fuzzy string matching and a weighted algorithm, all configurable via JSON.

Balcarek used Google AI Studio for drafting and agent mode for code completions, calling the process "vibe-coded." He found Google Cloud Platform approachable after initial debugging hurdles, like managing billing and API keys. The experience improved his prompt-engineering skills, creating reusable prompts for future projects and demonstrating a practical integration of AI into a web application.