Pinned Loading

1. Detecting_Fradulent_job Detecting_Fradulent_job Public

   Screening fraudulent job postings under a review budget: SQL rules (DuckDB) vs TF-IDF + LightGBM, evaluated on unseen posters

   Jupyter Notebook 1

2. Agentic_recommendation Agentic_recommendation Public

   Agentic RAG service: FastAPI + ChromaDB + LangChain, with a reproducible benchmark measuring a 42% embedding-latency reduction

   Python

3. BI_Dashboarding BI_Dashboarding Public

   Credit union member-services ops dashboard: PySpark/Spark SQL medallion pipeline, SLA & CSAT metrics, Databricks + Power BI/Tableau marts

   Jupyter Notebook

4. kafka-databricks-nasa kafka-databricks-nasa Public

   Streaming NASA gamma-ray burst alerts from Kafka into a Databricks medallion lakehouse — unit-tested notice parser, snowflake dimensional model

   Python

5. payments-data-platform payments-data-platform Public

   Payments platform: Postgres CDC via Debezium/Kafka into an Iceberg lakehouse served by FastAPI, plus a Snowflake/dbt batch warehouse normalising every payment to USD

   Python

6. medicare-claims-analytics medicare-claims-analytics Public

   Medicare claims analytics on CMS DE-SynPUF: dbt + DuckDB pipeline, PMPM & 30-day readmissions, high-cost member targeting, Streamlit + Power BI

   Jupyter Notebook