<- back to projects
Reversed ATS Platform (RATS)
RATS (Reversed ATS Platform) is a job-market intelligence platform that tracks live LinkedIn listings, visualizes hiring trends, and ranks open jobs against an uploaded CV.
What It Does
- Near real-time job dashboard: refreshed every two hours with filters, posting trends, and top hiring companies.
- CV-to-job matching: upload a PDF, TXT, or Markdown CV and get the closest live listings ranked by embedding similarity.
- Mobile-first experience: the dashboard, CV matching, and job detail views are built for phone screens.
Job Market Dashboard
CV Matching
Architecture
The pipeline moves data from a LinkedIn crawler into Confluent Cloud Kafka, processes it with Spark Structured Streaming on Databricks, stores a bronze Delta table, models it with dbt into an analytical Iceberg layer, materializes embeddings into Qdrant, and serves search and matching through FastAPI and Next.js.
Tech Stack
- Ingestion: LinkedIn crawler, Confluent Cloud Kafka
- Processing: PySpark and Spark Structured Streaming on Databricks, Delta Lake
- Transformation: dbt, Apache Iceberg
- Search: Qdrant vector database with embedding similarity
- Serving: FastAPI, Next.js
- Operations: GitHub Actions
Roadmap
- Salary prediction from skills, location, and experience
- NER-based tech-stack extraction from job descriptions