• By NeoMoti Team
• 24th July, 2026

Introducing Hojiinet: A Next-Gen Full-Stack Job Search Engine

Fragmented data and basic keyword searches are major challenges in regional job markets. Job seekers frequently run into irrelevant search results, while crawling and indexing systems often face scale issues. To address this, we developed Hojiinet: a modern, end-to-end full-stack platform built from the ground up to redefine job matching.

The architecture consists of a highly responsive Next.js frontend, styled with shadcn/ui and TailwindCSS, combined with a robust FastAPI backend. Under the hood, Hojiinet utilizes a custom, Tor-proxy-rotated crawling framework that aggregates vacancies from multiple portals securely and reliably without trigger-blocking issues.

Data ingestion triggers instant synchronization events. Leveraging RabbitMQ as an event broker, the system processes updates in real time. We integrated a 384-dimensional vector embedding model (all-MiniLM-L6-v2) alongside Meilisearch's hybrid search engine, matching candidates to job opportunities based on both precise keyword lookups and semantic intent.

By combining modern frontend patterns with asynchronous event-driven pipelines, Hojiinet delivers an exceptionally fast, fully localized, and responsive job-finding experience. It serves as a flagship demonstration of NeoMoti's capabilities in building real-world, scalable AI & full-stack applications.