Hassan Tavakoli

Architected and deployed a Data Lake on the GCP, integrating MongoDB and SDK. Collaborated with teams to review backend code and discuss code conventions, enhancing business logic understanding of data modeling. Reduced expert API execution time from 5 to under 1 minute by direct queries on BigQuery Analyzed data zone transformations, implementing Google DLP for enhanced data security and privacy. Deployed Kubernetes-orchestrated ETL/ELT via Airflow, cutting processing costs by $300,000 annually.

Configured Cloudera Apache distribution across 40 nodes and 2PT storage, leveraging over 5000 processor cores. Executed real-time data streaming with PySpark to calculate key decisions in under 1 minute. Deployed Apache technologies to prevent fraud by integrating extensive daily operational data from various sources. Fostered seamless collaboration by resolving issues with assertive communication and thorough documentation. Led 5 team members, closely collaborating with diverse teams to design.

Spearheaded microservices architecture with Kubernetes for near real-time bank card fraud detection. Utilized PySpark scripts for ad hoc analysis on 100 TB datasets stored in HDFS. Deployed Data Mart and cube with Kylin, reducing runtime from 15 to under 5 minutes. Effectively communicated with the core bank team to design the pipeline.