|
DATA SCIENTIST
PHOTOGRAPHER
TABLE TENNIS PLAYER
Story About Me
Chapter 1: The Spark
Hey! I’m Ammad— a Data Scientist, Photographer, and Tech Enthusiast. My fascination with numbers started in high school, where I excelled in Physics, Chemistry, and Mathematics. Solving equations and exploring analytical challenges shaped my journey toward engineering.
Chapter 2: Discovering Coding
During my Bachelor's in Mechanical Engineering degree, I discovered my passion for coding while designing a 4 Degree of Freedom (DOF) robotic manipulator in MATLAB. This manipulator, with its unique half-doughnut task space, introduced me to the world of programming and sparked a newfound curiosity.
Chapter 3: Reaching New Heights
My final-year project involved building a sailplane that generated electricity while flying at 100 meters. Programming the servo motors for the ailerons, flaps, rudder and elevator, using Arduino, was the highlight of the project, solidifying my love for technology-driven innovations.
Chapter 4: The Data Science Path
MIT’s Applied Data Science program ignited my passion for data science, helping me explore fundamentals of Python, Statistics, Machine Learning and Recommendation algorithms, and inspiring me to pursue a Master’s in Management Analytics at McGill University. At McGill, I deepened my expertise in tools like R, SQL, PySpark, Microsoft Azure, AWS and PowerBI, blending technical skills with business insights. These experiences prepared me to solve complex problems and make data-driven decisions with impact.
My Data Journey 🚀
Certification in Applied Data Science Program
MIT Professional Education
- Completed a capstone project on predicting used car prices using machine learning models, achieving 70% accuracy with Ridge regularization
- Predicted house prices in Boston and hotel booking cancellations using techniques like Linear Regression, Logistic Regression, and kNN
- Analyzed diabetes data from the Pima Indians tribe, focusing on exploratory data analysis, statistics, and visualization.
Started Master's in Management Analytics
McGill University
- Gained hands-on experience in Generative AI by becoming a finalist in the Retail GenAI Hackathon, where I developed a shoe-fitting chatbot solution using Llama 2 (70b parameters) and AWS
- Gained proficiency in querying languages like SQL and optimization libraries such as Gurobipy, applying them to solve complex analytical problems
- Leveraged multivariate statistics using R to analyze patterns and make data-driven decisions
- Integrated multiple data sources through APIs using Python to develop "Pillpal," a personalized customer support chatbot for a pharmacy, enhancing user experience and showcasing practical data integration skills
Data Engineer - Analytics Consulting Project
AtkinsRéalis
- Conceptualized and engineered three data ingestion and ETL pipelines using Azure Data Factory that consolidated data from 3 different sources, processing over 1 million records
- Optimized the medallion architecture by incorporating parallel processing and advanced SQL tuning, cutting data processing time by 27% (from 55 minutes to 40 minutes) and ensuring near real-time data availability
- Built a dynamic, metadata-driven pipeline that employed “Wait”, “Until” and “IfCondition” activities for real-time validation and error handling, enabling automatic reprocessing of incomplete data and ensuring high data reliability
Data Science Intern
Cineplex Entertainment
- Migrated and optimized LightGBM ML model scripts from Microsoft Azure ML to Databricks using Spark and Python, establishing a scalable data pipeline that improved runtime performance by 20%
- Developed a unified ETL workflow in Azure Databricks using PySpark to merge and transform the CSV files based on a defined schema, filtering out anomalies and ensuring accurate forecasts for both released and upcoming films under 8 minutes
- Leveraged the Hive metastore to implement a feature store in Databricks using Unity Catalog, preserving engineered features and metadata for consistent usage across multiple ML models and downstream applications
- Utilized Spark SQL temporary views to execute complex aggregations and load refined datasets from over 500 tables (each with 1+ million records), cutting data processing latency from minutes to seconds and enhancing overall pipeline throughput by 40%
- Maintained version control using Azure DevOps (Repos and Git), ensuring efficient tracking, management, and implementation of changes, fostering a streamlined development process
- Collaborated with BI, Data Science, and Data Engineering teams to design and integrate data solutions aligned with business goals, enhancing decision-making through actionable insights
Research Assistant - Prof. Galperin
McGill University
- Developed a scalable workflow using OpenAI's text-embedding-ada-002 model to process over 1 million rows, optimizing performance for large-scale datasets
- Automated data preprocessing and API integration with Python, converting large dataframes to JSONL files and consolidating outputs into a single structured output for analysis
- Optimized large-scale data processing using Dask DataFrame, enabling parallelized batch processing for enhanced pipeline performance
Finished Master's in Management Analytics
McGill University
- Explored Natural Language Processing (NLP) techniques, including lemmatization, stemming, sentiment analysis (VADER, BERT), and topic modeling (LDA), for advanced text analytics
- Developed a soccer player influence scoring model using centrality metrics and logistic regression, quantifying player impact within team passing networks for enhanced game analysis
- Developed an advanced, interactive Power BI dashboard for 24Seven's store expansion strategy, integrating historical sales, customer demographics, product trends, traffic flow, and economic indicators to optimize site selection (Calgary, Montreal, and Toronto) and forecast revenue growth
- Applied causal inference to analyze the impact of lead times on hotel cancellations, providing actionable insights for hospitality management
Featured Works
In Progress
In Progress
Tech Stack
Say Hello!
Montreal, Canada
+1 (438) 466-4396
ammadsohail2017@gmail.com