Project ideas that actually get noticed
Recruiter-facing projects now combine AI, cloud and a clear product arc — examples pushed this week include an AI-powered Chrome extension agent for smarter browsing, an AI interview-prep platform that scores code, and real-time NBA analytics dashboards that layer live stats and visualization. These projects pair modern stacks (Next.js, FastAPI, serverless) with demonstrable architecture and deployment details for portfolio impact. (x.com)(x.com)
A set of public GitHub repos this week shows recruiter-facing projects with explicit stacks: a voice-enabled AI recruiter built on Next.js, React, Vapi.ai and Supabase (tauhid017/-AI-Recruiter-Voice-Agent). (github.com (github.com)) An open-source AI interview-prep repo documents features like posture analysis, eye-contact scoring and resume ATS feedback under an AI Interview Prep project. (github.com (github.com)) A real-time NBA analytics repo pairs live game ingestion with ML moneyline predictions (logistic regression) and a signup/picks tracking flow in its README. (github.com (github.com)) Chrome-extension agent tooling in the wild includes an “Agentic Browser” with ~1,000 users on the Chrome Web Store, a multi-model automation extension (HARPA AI) that integrates ChatGPT/Claude/Gemini, and side‑panel workflow tools like BrowseWiz for in‑browser AI automations. (chrome.google.com (chromewebstore.google.com)) (harpa.ai (harpa.ai)) (browsewiz.com (browsewiz.com)) Modern full‑stack templates make the architecture and deployment steps reproducible: Next.js + FastAPI templates advertise end‑to‑end type safety (Zod + TypeScript), Docker and Vercel‑ready configs, and Vercel’s Next.js+FastAPI starter maps FastAPI to Python serverless functions in production. (nextfastapi.com (nextfastapi.com)) (vercel.com (vercel.com)) Live NBA dashboards cited by these projects commonly rely on packages and APIs for play‑by‑play and live box scores—examples include the community nba_api Python client, the data.nba.com/OpenAPI surface, pbpstats play‑by‑play parsers, and third‑party feeds like balldontlie for historical/live data. (pypi.org (pypi.org)) (sgsdm.github.io (sgsdm.github.io)) (pbpstats.readthedocs.io (pbpstats.readthedocs.io)) (balldontlie.io (balldontlie.io)) Recruiter‑facing guidance in hiring‑market writeups and GitHub community threads consistently flags what to show: 3–5 deployed projects with live URLs, clear README + architecture diagrams, CI/test evidence and measurable metrics (stars/users/uplift), because many hiring managers still inspect GitHub during screening. (tietalent.com (tietalent.com)) (github.com (github.com)) (techtimes.com (techtimes.com))
