AbdulHafeez S. | BI & Analytics Engineer | AI-Augmented SWE | AWS, dbt, Python | Building Agentic Data Products

AbdulHafeez S.

BI & Analytics Engineer | AI-Augmented SWE | AWS, dbt, Python | Building Agentic Data Products

Based in Malaysia

About

BI & Analytics Engineer with a unique foundation in mechanical, CAE, AI & software engineering. I specialize in building decision-ready data products by combining deep analytics expertise with AI-accelerated software engineering. My background allows me to approach analytics with a product mindset, utilizing agentic workflows to rapidly prototype, test, and deploy automated data pipelines on AWS and GCP. I leverage AI tools to enhance every stage of the lifecycle: from automated schema design and dbt modeling to generating interactive dashboards via data viz tools. This results in faster iteration cycles and higher-quality data narratives. Driven by empowering stakeholders with clear, actionable insights, Eager to apply my blend of agentic engineering to solve complex business problems.

Skills

Projects

InsightFlow - AWS Retail & Economic Data Pipeline

Built an automated AWS data pipeline using Terraform, Kestra, and Python to analyze Malaysian retail performance vs. fuel prices. Processed raw S3 data into a star schema with dbt, enabling analytics via AWS Athena and an interactive QuickSight dashboard.

Divvy Bikes - AWS Data Engineering & BI Platform

Architected a production-ready business intelligence platform on AWS to drive marketing strategy for converting casual bike-share riders into members. Managed cloud resources with Terraform and built an automated ETL pipeline using Docker-based Airflow processing 11M+ records in Redshift.

Data Career Navigator - Job Market Analytics Platform

Developed a data platform for LinkedIn job insights: automated ETL with Python and Medallion architecture, used NLP to extract skills/salary data, orchestrated with Airflow, and delivered analytics via MotherDuck and a Streamlit dashboard.

BI Outlier Detection System

BI-focused anomaly detection with calibrated uncertainty. Built with Python, scikit-learn, and Streamlit. Reproducible notebooks show techniques; a live dashboard lets analysts explore anomalies, rank risk by confidence, and reduce false positives.

Building an Interactive Sales Dashboard for Data-driven Decision-Making at Cycling Superstores

Cycling Superstores lacked a centralized & user-friendly platform to analyze sales data. This made it difficult for executives to gain insights into sales performance across various metrics like product category, brand, store location, and customer demographics. Hence executive dashboard is a must.