fullstack

Orchid Island - AI-Powered ATS & Microservices Recruitment Platform

An enterprise-grade, event-driven recruitment platform designed to automate end-to-end talent acquisition. Built with 6 Java Spring Boot microservices, a Python FastAPI AI/NLP service, Next.js 14 frontend, MinIO S3, and RabbitMQ, it features privacy-preserving local LLM CV parsing and objective mathematical ATS candidate ranking.

Impact

Eliminated manual resume screening bottlenecks with >80% time reduction, ensured 100% candidate data privacy through local Ollama inference, and unified the hiring lifecycle from application to ERP onboarding.

Key Highlights

Engineered 6 Spring Boot microservices (API Gateway, Auth with RBAC, Recruitment, ATS Scoring, Interview Lifecycle, Notifications) orchestrated via Docker Compose.
Integrated local LLM inference via Ollama (qwen2.5:7b) and fallback providers (Gemini, OpenAI) for zero-data-leakage CV extraction with Pydantic validation.
Developed a weighted mathematical ATS scoring engine evaluating skills, experience, education, and language match with customizable department weights.
Implemented event-driven asynchronous processing via RabbitMQ and MinIO S3 for high-throughput batch CV ingestion and email notifications.
Integrated bidirectional Calendly webhooks for automated interview scheduling and Collabix ERP integration for hired candidates.

Tech Stack

JavaSpring BootSpring Cloud GatewayPythonFastAPINext.jsTypeScriptTailwind CSSRabbitMQPostgreSQLDockerDocker ComposeMinIOOllamaPydanticPyMuPDF