Arslan Aslam

I

AI-Powered E-Commerce Analytics Engine with RAG and Agentic Workflows

Independent project

Apr 2026 - May 2026 (2 months)

• Engineered a Medallion Architecture pipeline (Bronze → Silver → Gold) orchestrated by Apache Airflow, transforming 100K+ raw Olist e-commerce records into a star schema with fact and dimension tables in PostgreSQL • Built a RAG pipeline using Voyage AI embeddings and ChromaDB, enabling semantic search over 40K+ customer reviews for qualitative business insights with sub-second query response times • Designed an agentic LangGraph workflow with intelligent query routing — automatically classifying user questions and directing them to SQL, RAG, or hybrid retrieval paths for comprehensive answers • Containerized the full stack (FastAPI, Streamlit, Airflow, PostgreSQL, ChromaDB) using Docker Compose, enabling one-command deployment and full re

I

Real-Time Smart City Data Streaming Pipeline on AWS

Independent project

Jan 2026 - Feb 2026 (2 months)

• Architected a real-time streaming pipeline using Apache Kafka and Apache Airflow to ingest and process high-volume sensor data from 5+ smart city data streams • Leveraged AWS services (S3, Glue, Redshift) to build a scalable lakehouse-style storage layer supporting transformation and analytical querying • Containerized all pipeline components with Docker and implemented automated monitoring and error-handling logic, achieving 99%+ pipeline uptime

I

End-to-End AWS ETL Pipeline with Apache Airflow

Independent project

Oct 2025 - Nov 2025 (2 months)

⚙️ Workflow Sequence: 🔹 The project begins with Data Cleaning & Validation of raw CSV files using an AWS Glue ETL Job Notebook, ensuring the datasets are consistent and analysis-ready. 🔹 The cleaned files were then stored securely in Amazon S3, serving as the central data lake for the pipeline. 🔹 Next, I created an AWS Glue Database and configured Crawlers to automatically infer schema and build data catalogs for the cleaned datasets — making them queryable in AWS Athena. 🔹 Using Amazon Athena, I wrote multiple SQL queries to extract valuable insights and perform analytical operations such as data filtering, aggregation, and optimization directly on S3 data. 🔹 For workflow automation, I integrated Apache Airflow locally through Do

I

Sales ETL Pipeline

Independent project

Oct 2025 - Nov 2025 (2 months)

Objective: To design a modular ETL (Extract, Transform, Load) process for sales data using modern data engineering tools and practices. 🔹 Tools & Technologies Used: 🐳 Docker – to containerize Apache Airflow and PostgreSQL for smooth local orchestration. ⚙ Apache Airflow – to schedule and automate ETL workflows through DAGs with task dependencies and custom logging. 🐘 PostgreSQL (via DBeaver) – to perform data cleaning, filtering, and transformations using modular SQL scripts. 💻 Python (with psycopg2, Pandas, Jupyter Notebook) – for connecting to the database, validating data, and handling transformations programmatically. 🔹 Workflow Sequence: Extracted raw sales data and loaded it into PostgreSQL. Performed data cleaning, filtering

I

ETL Pipeline

Independent project

Sep 2025

🔹 What this project does: Extracts raw transaction data into Databricks Transforms the data with PySpark & SQL (cleaning, aggregation, filtering) Loads structured results into tables for business insights 🔹 Skills applied: Databricks | Apache Spark | PySpark | SQL | Data Engineering Concepts

I

Online Voting System

Independent project

Dec 2024 - Jan 2025 (2 months)

A secure, province-based voting simulation implementing OOP and data management. ✔ Key Features: User Authentication: Login/registration system with password strength validation and unique voter ID checks. Province-Based Voting: Voters can only vote for candidates from their registered province (Punjab, Sindh, Balochistan, KPK). Admin Controls: View voter records, candidate vote counts, and election results. Data Integrity: Input validation to prevent invalid votes or duplicate registrations. Vote Tracking: Real-time vote counting and result declaration per province. ✔ Technical Highlights: OOP Concepts: Used struct to manage voter data (ID, province, votes) and modular functions for each operation. Memory Management: Employed pointers to

About

Detail

Timeline

Résumé