Hello, I'm
Imran Ahmed

APPLIED AI & DATA SCIENCE | SOFTWARE DEVELOPER

PROJECTS

OLap project

Semantic Layer & OLAP Cube Project

Nov 2025 - Dec 2025

Designed and implemented a semantic layer for e-commerce data with consistent business metrics (Revenue, Order Count, AVG Order Value). Demonstrated OLAP operations and pre-aggregation performance benefits, showcasing efficient data warehousing and analytics patterns.

Tech Stack: Metabase, Docker, Sample Database

Learn More
Azure Data Factory ETL Pipeline project

Automated Sales Data ETL Pipeline using Azure Data Factory

June 2025 - Aug 2025

University of Windsor
Developed a production-ready Azure Data Factory ETL pipeline integrating Blob Storage, Data Flows, and SQL Database. Transforms raw sales data into analytics-ready datasets with data filtering, derived columns, and schema mapping. Automated daily pipeline execution using ADF triggers with parameterized inputs and source-controlled deployment via GitHub.

Tech Stack: Azure Data Factory, Azure Blob Storage, Azure SQL Database, Azure Data Flows, SQL, Git, GitHub, Cloud Automation

Learn More
Intelligent Document Q&A System with LLMs and RAG for Enterprise Knowledge Management

Intelligent Document Q&A System with LLMs and RAG for Enterprise Knowledge Management

May 2025 - June 2025

Built document ingestion and ETL pipelines feeding a data lake-backed AI application. Implemented semantic search and monitoring for reliable, low-latency query responses.

Tech Stack: Python, Azure, LangChain, ChromaDB, Docker

Learn More
PointRide ride-sharing platform project

PointRide – A Secure Multi-Role Ride-Sharing Platform

May 2025 - Aug 2025

A full-stack Django and Bootstrap ride-sharing platform with multi-role authentication, dynamic booking with seat management, and Google Maps integration. Features secure role-based dashboards for Travelers, Drivers, and Admins, with comprehensive data validation and responsive design for seamless cross-device experience.

Tech Stack: Django, Bootstrap, SQLite, Google Maps API, JavaScript, Git, Python, HTML/CSS

Learn More
Adaptive Query Optimization project

Adaptive Query Optimization through Dynamic and Predictive Sharding in Large-Scale Databases

Jan 2025 - Apr 2025

University of Windsor
Dynamic Sharding adaptively redistributes data based on real-time workload patterns, while Predictive Sharding employs machine learning models to anticipate future data growth and optimize shard allocation proactively. The proposed methodology is evaluated using a synthetic dataset, leveraging Linear Regression, Logistic Regression, Random Forest, and Artificial Neural Networks to predict execution time and guide sharding decisions.

Tech Stack: Python, SQL, Machine Learning, Data Sharding, Linear Regression, Logistic Regression, Random Forest, Artificial Neural Networks

Learn More
Movie Rating Prediction project

Movie Rating Prediction Using Transformer-Based Models

Jan 2025 - Apr 2025

University of Windsor
Developed a transformer-based approach utilizing the BART model to predict movie ratings from plot summaries. Compared BART against traditional ML models (Naïve Bayes, SVM, Logistic Regression, KNN, Random Forests, AdaBoost) and BERT. BART achieved an F1-score of 0.81, significantly outperforming all baselines, demonstrating the superiority of contextualized embeddings for text-based rating prediction.

Tech Stack: BART, Transformers, BERT, Naïve Bayes, SVM, Logistic Regression, KNN, Random Forests, AdaBoost, Python, NLP

Learn More
Vehicle data analysis project

Vehicle Data Analysis and Predictive Modeling Using R

Jan 2025

This project focuses on comprehensive vehicle data analysis using the R programming language and R Studio, showcasing various techniques in data preprocessing, exploratory data analysis (EDA), and data visualization. The goal was to explore a vehicle dataset containing details like car prices, fuel types, transmission types, and other vehicle characteristics, enabling insightful analysis that can assist in pricing prediction and understanding vehicle market trends.

Learn More
GNN project img

Research Project: Mitigating Racial Bias in Natural Language Understanding with Graph Neural Networks

Sept 2024 - Dec 2024

Natural Language Processing (NLP) models often inherit racial biases present in textual data. By capturing semantic relationships with text, this research investigates the use of Graph Neural Networks (GNNs) to address racial bias in language understanding. Our approach shows how GNNs can be used to mitigate bias and fairly represent minority groups in NLP tasks.

Learn More
Mobile app project img

Pick a Plan - Mobile Application

Sept 2024 - Dec 2024

The PickaPlan project aims to create a Java-based software application for analyzing mobile data plans. By leveraging advanced data structures and algorithms, the project focuses on providing accurate and efficient analysis of data plans based on critical features such as cost, data allowance, overcharge conditions, and plan benefits. This initiative integrates concepts learned in class to develop a practical, real-world application.

Learn More
Customer classification project img

Customer Classification based on Car Purchase (WEKA Project)

Aug 2022

The aim of this project was to implement the 2 different supervised learning algorithms on a labelled dataset and an unsupervised learning algorithm on an unlabeled dataset. Weka tools had been used for developing the models and analyze the results and discuss the effectiveness of the algorithms for that specific dataset.

Thesis img

Academic Thesis: Comparison of Moving Average Model and Deep Learning Model in Terms of Stock Price Prediction

Sep 2021 - Apr 2022

The stock market's unpredictability and complexity make predicting stock prices challenging. Despite this, financial experts use algorithms, including machine learning and deep learning, to identify patterns. This research compares traditional technical analysis models (SMA and EMA) with deep learning (LSTM) for predicting stock closing prices.

Learn More
Handwritten classification project img

Research Project: Bangla Handwritten Digit Classification Using Raw Image Data

Feb 2022 - Apr 2022

In this project, the main focus was to classify Bangla handwritten digits from raw image data. To accomplish this project, a dataset has been made of raw images and a convolutional neural network is also constructed to classify the images. An Alexnet approach (mean value subtraction) was used to normalize the raw images.

Learn More
TalkBuzz project img

TalkBuzz: Web Based Real-Time Chatting Application

Jan 2021 - Apr 2021

The aim of this project is to create a web-based chatting application which will allow users to communicate in real time using easily accessible web interfaces. It will be a type of online chat, distinguished by its simplicity and accessibility to the users who do not wish to take the time to install and learn to use specialized chat software. It is an application which will help to communicate with friend using internet. Users have to register and obtain credentials to login into the application.

Learn More
covid test management system img

COVID-19 Test Management System Project (Object Oriented Analysis & Design Approach)

Aug 2020

The system allows users to securely log in with a username and password, offering features like account recovery, optional email registration, and password retrieval via a verification code. Users can apply for COVID-19 testing, share their location via GPS, make payments, and receive test results. The system prioritizes usability, security, and performance, ensuring reliable and efficient operation with high availability and scalability.

Transit management system img

Transit Management System using Object-Oriented Analysis & Design Approach

Mar 2020 - May 2020

The system consists of various components: users, drivers, bus companies, GPS, and QR codes. Users create accounts to search for buses, view routes, stops, and costs, and track buses via real-time GPS on Google Maps. Drivers and bus companies also have accounts, enabling them to manage routes, costs, and notices. QR codes on buses provide detailed information, while the app facilitates communication and feedback between passengers, drivers, and bus companies, ensuring a seamless experience.

Car Rental Projct img

Car Rental Management System Database

May 2019 - Aug 2019

The primary objective of this system is to offer an efficient and easily manageable data storage and modification facility. The storage design has been developed based on the specific requirements of the project. This system is intended to enable administrators, managers, and employees to manage and update relevant database information across various fields. Additionally, the administrator will have the authority to add or remove employee information as required.

PROFESSIONAL EXPERIENCE

Data Scientist (Master's Co-op)

Sept 2025 - Dec 2025

Exeevo, Toronto, Ontario, Canada

  • Designed and deployed an LLM-powered AI voice assistant leveraging Azure Cognitive Services and Retrieval-Augmented Generation (RAG) to enable secure, context-aware enterprise workflows.
  • Delivered production-grade machine learning and LLM solutions supporting automation and intelligent decision-making across internal teams.
  • Built and validated data pipelines, integrated REST APIs, and supported containerized deployment using CI/CD workflows for scalable systems.

AI Engineer Intern

Oct 2024 - Aug 2025

MealLens AI, Windsor, Ontario, Canada

  • Collaborated on the design of a large-scale computer vision pipeline for automated meal recognition using a 200GB+ dataset.
  • Implemented annotation and feature representation techniques with Segment Anything Model (SAM/SAM2) and OpenAI CLIP to enhance accuracy for meal and ingredient labeling.

Junior Data Engineer

Jan 2023 - Aug 2024

Hydroque, Dhaka, Bangladesh

  • Designed and optimized scalable ETL pipelines processing over 1M annual water quality data points.
  • Developed real-time Tableau dashboards to deliver operational analytics and visual insights for stakeholders.

Student Intern

May 2022 - Aug 2022

American International University Bangladesh

  • Assisted faculty members in preparing learning materials and attended to student queries of the assigned faculty member.
  • Helped the assigned faculty member in research works and tracked course progress-assessment dates, deadlines, etc.

TESTIMONIALS

"I had the opportunity to work with Imran as part of MealLens, where he contributed to several technical projects related to AI and food recognition. During our time working together, Imran supported tasks such as dataset preparation, image annotation, and some aspects of computer vision and language integration. He approached his work with enthusiasm, was receptive to feedback, and communicated clearly with the team. Imran showed a good attitude toward learning and adapting in a collaborative environment. I believe he has the potential to grow in a technical role and contribute effectively to a team."

— Farzin Niaki, PhD

R&D Director, MealLens Inc.

"During the internship program he has been exposed to different areas of the function and displayed a commitment to learn and contribute. We wish him the best for all future endeavours and a successful career."

— Derek Evans

Chief Executive Officer, Exeevo

"Imran has been instrumental for the development of our data infrastructure and our ML models, he is a kind and enthusiastic individual that always delivers on time. What I like about Imran is that even with a new task he is keen to get started and understand and delve deeper into the task given."

— Qi Zao

Lead Data Scientist, Exeevo

"Imran is a highly motivated learner who consistently delivers work of exceptional quality. He takes the time to thoroughly understand the context of the sectors he contributes to, such as pharmaceutical sales, and approaches each task with a strong commitment to excellence. His dedication and attention to detail have made him an invaluable member of our team."

— Paven Rai

Data Scientist, Exeevo

REPORTS & PUBLICATIONS

Movie Plot dataset

Dataset Publish: 7k+ Movie Plot Dataset

This dataset contains information about movies gathered from IMDB and other sources. It includes the following features: Title: 7k+ Movie Plot Dataset. Plot: A brief summary or description of the movie's story. Genres: The genres or categories the movie belongs to (e.g., Drama, Action, Comedy). Countries: The countries where the movie was produced. Languages: The primary languages spoken in the movie. Average Rating: The average user rating given to the movie. Number of Votes: The total number of user votes or reviews the movie has received. The data has been cleaned and preprocessed to remove unnecessary symbols and text, providing a more streamlined and usable version for analysis.

Dataset DOI
Bangla handwritten dataset

Dataset Publish: Bangla Handwritten Digits Dataset (raw image)

This dataset contains 200 Bangla handwritten digit images. All the digits are handwritten on white paper by the author then the images are taken using a smartphone camera. After taking the images extra white areas are cropped.

Dataset DOI
Activation function img

Academic Report: The Impact of Activation Function in Neural Network

In this paper, various existing activation functions and its impact in deep learning has been discussed, also provided a brief overview about the requirement for activation function and non-linearity in deployment of deep neural networks. Many forms of activation functions and their effectiveness are also discussed. Suitable activation function for specific application is an important matter, and this collecting’s will help making effective decision in case of choosing suitable activation function for specific application.

Bangla handwritten dataset

Academic Report: Less Skilled People in Internet Service Providing in Bangladesh

The purpose of this report is to present a solution to remove the internet connectivity problem in broadband internet in Bangladesh. To solve this problem, internet service provider (ISP) should open their training programs for their employees. Some government organizations can also take some initiative to remove this problem. A survey has been conducted to collect users' opinions from broadband internet users. After analyzing the collected data, it is found that most broadband users are suffering from internet connectivity problem and agrees with the proposed solutions.

ACADEMICS

Master of Applied Computing (Artificial Intelligence Stream)

September 2024 - December 2025

University of Windsor

Windsor, Ontario, Canada

Bachelor of Science in Computer Science and Engineering

May 2018 - May 2022

American International University Bangladesh

Dhaka, Bangladesh

COURSE & CERTIFICATION

Sterling Cybersecurity

Sterling’s Cybersecurity Workshop Ethical Hacker, Network Defense, & Digital Forensics

Sterling Information Technologies inc, University of Windsor, EC-Council

Azure Certificate

Getting Started with Microsoft Azure

LinkedIn Learning

Hadoop Certificate

Learning Hadoop

LinkedIn Learning

Demo day 2024 Certificate

Certificate of Participation in the 11th Annual School of Computer Science Demo Day

Univeristy of Windsor

Unix Certificate

Unix Essential Training

LinkedIn Learning

Prompt Engineering

Introduction to Prompt Engineering for Generative AI

LinkedIn Learning

Python specialzation

Python For Everybody Specialization from University of Michigan

Coursera

Deep learn and computer vision

Deep Learning and Computer Vision A-Z: OpenCV, SSD & GANs

Udemy

Machine Learning

Machine Learning

Sololearn

AI for everyone

AI For Everyone

Coursera

reinforcement learning with python

Reinforcement Learning with Python

Great Learning Academy

aws fundamentals

AWS Fundamentals

Coursera

Data Science Foundation

Data Science Foundations

Great Learning Academy

data visualization using tableau

Data Visualization using Tableau

Great Learning Academy

image denoising

Image Denoising Using AutoEncoders in Keras and Python

Coursera Project Network

movie recomendation project

Movie Recommendation System using Collaborative Filtering

Coursera Project Network

cybersecurity badge

Introduction to Cybersecurity Tools & Cyber Attacks

Coursera

big goals

Making Big Goals Achievable

LinkedIn Learning

big goals

Selenium Essential Training

LinkedIn Learning

WORKSHOPS

Sterling’s Cybersecurity Workshop Series

Sterling’s workshops on Cybersecurity, hosted by Sterling Information Technologies Inc. in collaboration with the University of Windsor and EC-Council. This workshop series provided me with valuable insights and hands-on experience in the field of cybersecurity. The workshops covered a range of topics, including ethical hacking, network defense, and digital forensics.

Learn More

Technical Workshop - New Technology Trends in GenAI

In this workshop, Guest Speaker Naresh R. Jasotani, Principal Technologist at Google, provided an in-depth analysis of emerging trends in Generative AI (GenAI) and projected future advancements in the field. Mr. Jasotani elaborated on Google's large language model, Gemini, and introduced tools such as NotebookLM, GenAI Playground, and Gemini Studio. He emphasized the transformative potential of GenAI agents and forecasted that the year 2025 would be a critical period for their development and widespread implementation. The workshop offered valuable insights into the evolving landscape of GenAI, highlighting its implications for various industries and the innovative applications that are expected to emerge.

Learn More

Technical Workshop - Unlocking the Power of Recommender Systems by: Bahareh Rahmatikargar

This workshop marks the second workshop in a three-part series dedicated to the fascinating field of social network analysis and its application in recommender systems. Step right into the fascinating world of recommender systems! Our previous workshop explored various recommender systems, from traditional models to the latest advancements. We discussed their strengths and limitations, the data they use, and how to measure their performance.

Learn More

Technical Workshop - Advanced Topics in Data Clustering by Ali Abbasi Tadi

Clustering is a way of grouping data points into different clusters consisting of similar data points. The objects with possible similarities remain in a group that has less or no similarities with another group. In this workshop, we explore advanced topics in data clustering for high dimensions. We will define the curse of dimensionality and dimensionality reduction methods like t-SNE, PCA, and Isomap, as well as their implementations. We will explore the consensus matrix and how to adjust the best clustering parameters.

Learn More

Technical Workshop - Basic Topics on Cloud Computing for IoT by Atefeh Gilvari

This workshop explores the integration of Cloud Computing with the Internet of Things (IoT), detailing how data from loT devices is managed, analyzed, and utilized for decision-making through platforms like AWS. It highlights the architecture of AWS for loT, including device connectivity, security, and the practical applications of cloud technologies in enhancing efficiency and reliability in IoT ecosystems.

Learn More

Technical Workshop - Foundations of Q-Learning by Xiaofeng (Michael) Liu

Reinforcement learning (RL) is training machine learning models to make a sequence of decisions. Q-learning is a type of RL with a model-free environment which achieves learning by interaction with the environment instead of transition probabilities. This workshop will introduce the principle of Q-learning.

Learn More

Technical Workshop - Unveiling the Potential of Deep Reinforcement Learning (DRL): Methods and Applications (1st) by Nahid Abdolrahmanpour

Deep Reinforcement Learning (DRL) has become a groundbreaking paradigm in artificial intelligence, exhibiting the potential to solve complex decision-making tasks. This workshop aims to provide participants with a complete understanding of the methods and applications of DRL, highlighting its significance in various domains, including robotics, gaming, and autonomous systems. Additionally, the workshop will delve into recent advancements and prospects, empowering participants to unlock the full capabilities of DRL and its transformative impact on the technological landscape.

Learn More

Technical Workshop - Azure Machine Learning (2nd Offering) by Mahshad Hashemi

In this workshop, participants will delve into the basics of Azure Machine Learning, exploring its comprehensive ecosystem and the role it plays in the Al and ML landscape. We will cover essential components, the machine learning workflow, and the tools available for efficient model development.

Learn More

Technical Workshop - Intro to NLP (2nd Offering) by Soroush

Natural Language Processing (NLP) is a rapidly growing field of Artificial Intelligence that focuses on enabling machines to understand, interpret, and generate human language. This introductory workshop will cover the basics of NLP, its applications, and popular techniques, including text classification, sentiment analysis, named entity recognition, and machine translation.

Learn More

Cultural Workshop - Iran: A Journey Through Culture and Tradition (2nd Offering) by: Nasrin Tavakoli, and Atefeh Gilvari

Embark on a journey through the vibrant cultural tapestry of Iran! This interactive workshop invites you to delve into the heart of Iran's traditions, spanning its music, dance, and the rich diversity of its ethnic communities. From the soul-stirring melodies of Persian classical music to the high-energy folk dances of regions across the country, you'll experience the profound beauty and variety of each cultural expression. We'll uncover Iran's colourful history, artistic heritage, and the unique contributions of its many ethnic groups-from the Persian and Azeri to the Kurdish and Baluchi-each adding a distinct flavour to Iran's cultural mosaic.

Learn More

MSc Thesis Defense: HyCAN: A Hybrid Approach for CAN Bus Intrusion Detection using performance analysis of Machine Learning Techniques.

Electronic Control Units (ECUs) have become important to modern vehicles, enhancing operational control, driving comfort, and safety. These ECUs communicate using the Controller Area Network (CAN) protocol, which despite its widespread adoption, is vulnerable to various security threats. Therefore making CAN protocol more secure is crucial, intrusion detection systems (IDS) offer a viable path to mitigating cyberattacks on vehicles.

Learn More

PhD Defense: Federated Learning: Taxonomy, Challenges, Future Directions, and a Brief Insight into Graph Neural Networks for Decentralized Collaboration

Deep Learning is one of the most revolutionary concepts in the field of Artificial Intelligence, allowing us to train a Machine Learning model for almost any type of problem using any type of data. Federated Learning (FL) is a type of distributed Deep Learning framework in which the model is trained locally on each device, and only the trained gradients, also known as “local updates”, are sent to a central server that aggregates them and creates a global model.

Learn More

EVENTS

wetech innovation img

WETech Alliance Innovation Blueprint at University of Windsor

August 7, 2025

WETech Alliance, in partnership with the University of Windsor, is thrilled to announce the launch of the Innovation Blueprint, a comprehensive program designed to foster innovation and entrepreneurship among students. The Innovation Blueprint will provide students with access to resources, mentorship, and opportunities to develop their ideas into viable businesses. We are excited to see the innovative solutions that will emerge from this program and look forward to supporting the next generation of entrepreneurs at the University of Windsor.

Learn More
12th Demo day img

12th CS Demo Day at Univeristy of Windsor

April 4, 2024

We had the incredible opportunity to showcase our research on Energy-Efficient Query Optimization in Data Warehouses at the University of Windsor’s CS Demo Day Winter 2025. A huge thank you to Dr. Shafaq Khan for her invaluable guidance and support throughout this journey.

Learn More
Starling cybersecurity last day

Final Day at Sterling Cybersecurity Workshop Series

March 26, 2024

In the final day at the workshop, participants are exposed to various cybersecurity challenges and are forced to think outside the box to find solutions. This helps to sharpen problem-solving skills, improve analytical thinking, and develop a deeper understanding of cybersecurity concepts.

Learn More
11th Demo day img

11th CS Demo Day at Univeristy of Windsor

November 5, 2024

Our project, PickaPlan, was showcased at the 11th CS Demo Day hosted by the School of Computer Science at the University of Windsor.

Learn More WETech Alliance
Podcast

Podcast on AI: Threat or Gateway to New Opportunities

October 26, 2024

This podcast explores the dual nature of artificial intelligence (AI) as both a potential threat and a gateway to new opportunities in the job market.

Special Guest: Md. Fahim Khan

Software Developer at Rocket Innovation Studio

GDG Devfest 2024

Google Developer Group (GDG) DevFest 2024 Windsor: Streamlining Cloud-Ready Applications with Docker and Google Cloud

November 16, 2024

This session, was about how to containerize applications and deploy them seamlessly to Google Cloud. Also, Docker’s role in packaging applications and Docker Compose’s ability to streamline local development by mimicking cloud environments.

Learn More
GDG Devfest 2024

Google Developer Group (GDG) DevFest 2024 Windsor: Implementing TARA for Automotive Cybersecurity: Insights and Best Practices

November 16, 2024

Threat Analysis and Risk Assessment (TARA) as the foundational stage of cybersecurity, which has been highlighted in key standards and best practices, such as MITRE, NIST, ISO/SAE 21434, and Auto-ISAC. The presentation is based on the growing importance of cybersecurity in automotive systems, particularly as vehicles become increasingly connected and automated.

Learn More
Podcast

Volunteering at Fall Open House 2024 at University of Windsor

October 26, 2024

As a volunteer, I had the opportunity to assist prospective students and their families in navigating the campus, answering questions, and providing information about the programs and services offered at the University of Windsor.

Podcast

Tech Connect Thursday | A Spotlight on Careers in Data Analytics by WeTech Alliance

October 3, 2024

Andrew Vendrasco, Data Analyst at Libro Credit Union, shared his insights on the field of data analytics. He discussed his career journey, the skills required for success in the field, and the impact of data analytics on various industries. The session also included a Q&A session where participants could ask questions and engage with the speaker.

Podcast

Windsor-Detroit’s Digital Twin Pt.III by Hackforge

September 9, 2024

Joined the third and final part of the Windsor-Detroit’s Digital Twin series, where the potential of the digital twins in urban planning, infrastructure development, and sustainability was discussed. The expert panel also discuss the challenges and opportunities of implementing digital twins in the Windsor-Detroit region, and the transformative impact they can have on community engagement, decision-making, and future development.

SKILLS

Python logo
Keras logo
Tensorflow logo
Pytorch logo
Pandas logo
C logo
C plus plus logo
C sharp logo
Java logo
MySQL logo
Git logo
Github logo
HTML logo
CSS logo
Javascript logo
PHP logo
Bootstrap logo
Tableau logo
Weka logo
Linux logo
Colab logo
Jupyter logo
VS code logo
Pycharm logo
Intellij logo
Android Studio logo
Matlab logo
Autocad logo
Jenkins logo
MLflow logo
Kubernetes logo
Docker logo
Imran's Img

About Me

Imran Ahmed, he/him, born in Dhaka, Bangladesh, completed his post graduation from the School of Computer Science, University of Windsor, Ontario, Canada. He spent his early childhood in KSA. He completed his Bachelor's degree in Computer Science and Engineering from American International Univeristy Bangladesh Majoring in Computational Theory.

Upon his graduation, he worked with his thesis supervisor in research works in AI and Machine Learning field.

He has gained valuable experience in programming, algorithms, data structures, and software development principles through his studies and various projects. Imran is particularly interested in using AI and machine learning to solve real-world problems and improve lives.

Imran is a proactive learner and is open to early career opportunities, willing to relocate and contribute to exciting advancements in the field.

LinkedIn Github

Get In Touch

My Contact Details

Email: imranahmed9813@gmail.com

Address: Toronto, Ontario, Canada