Hi, I'm Tazwar.
Welcome to my site!

Who? Contact me!

About Me

Originally from Dhaka (Bangladesh), I moved to Vancouver back in 2016 to pursue my undergraduate studies here at The University of British Columbia, where I am studying Statistics.
Coding during the day,playing the guitar out loud at night; that's essentially the dream life for me and always has been. Here's a closer look at two of the most essential components of my life ...

Music

Since the very day I saw my father shred a double-tapped guitar solo on national television back home in Bangladesh, I knew I wanted to learn how to do that. So, with his guidance I started out with the guitar back in 2010, and have been playing ever since then. I've even played onstage back home, with my father and NOTHING on Earth comes close to the feeling of shredding a solo in front of a screaming audience. In my spare time, I enjoy nothing more than playing my electric guitar and learning how to play classic tracks by my favourite bands (Van Halen, Judas Priest, Metallica and Megadeth just to name a few).

Music Image

Coding

I probably saw the first Iron Man movie back in 2012 (yes, after The Avengers). Until that point, I had never really had a clear vision of what I wanted to do in life. But watching Tony Stark working on the first suit just fascinated me. I wanted to do THAT. More precisely, I wanted to create something akin to J.A.R.V.I.S.. I wanted to learn about Artficial Intelligence from a practical perspective and I finally got around to learning Python in 2017.
Flash forward to the present day, after having learned the fundamentals of programming and tackling Data Science problems via coursework/books/online tutorials, I have worked 8 months as an Artificial Intelligence Researcher Intern with Huawei at the Big Data & Intelligence Platform Lab. I have also spent 8 months as a Machine Learning Researcher (Intern) with the National Research Council of Canada.

Code Image

Education

I started my undergraduate degree in the Engineering faculty and finished a year of Civil Engineering with Dean's Honours List standing. But, I switched out to Statistics in 2018, so that I could bite deep into the fundamental Mathematical theory in school and improve my practical programming skills on my own time.
I've listed some of the core courses I have taken so far which have helped me hone my abilities as a Software Developer and also prepared me very well for understanding the inner-workings of Machine Learning models. Feel free to hover any of them and click to check out the description on the official UBC site!

Statistics

Computer Science

Mathematics

Technical Knowledge

My first exposure to programming was in my first semester of university; and I was terrible at it. Diving headfirst into the language C felt like a nightmare for a beginner like me. But there was something so satisfying about coding that reminded me of the thrill of creating solos on the guitar. And since then, I have approached programming like an art, with the hunger to learn different concepts/languages and apply them to create things. Here is a list of components in my growing repertoire of knowledge.

Programming Languages
  • Python
  • Java
  • JavaScript
  • TypeScript
  • Golang
  • C++
  • R
  • HTML5/CSS
  • C
Libraries & Frameworks
Development Environments

Projects

The process of learning a particular concept naturally compels me to try and apply it to a practical problem. Here are some of the ways I have used my programming knowledge to either analyze real data, make an interactive application or in some other way improve my skills as a developer.

Whats-Hot Image

Whats-Hot?

Python, JavaScript

Find the most popular restaurants in Vancouver (BC) according to recent customer reviews, with 1 click! To develop this app, I trained an L1-regularized Logistic Regression classifier on 1 million Restaurant Reviews (85.6% accuracy) and linked it with a Flask-React app (designed by yours truly) to implement an application that sorts restaurants based on positivity in customer reviews.

Youtube Demo
Analyzer IMG

Stock Market Analyzer

Java

Implemented an interactive analyzer with real-time stock market data from the Alpha Vantage API, to investigate technical indicator crossovers and general market momentum of companies side-by-side, along with identifying potential up-trends/downtrends. Indicators included Simple and Exponential Moving Averages, MACD lines, etc.

Youtube Demo
Traffic IMG

Website Traffic Analysis

Python

Using a dataset on website visitors from the UCI Machine Learning repository, this project aims to answer the following question: will a website visitor with a specific set of characteristics end their browsing session with a purchase or not? With the aid of Principal Component Analysis, the final Naive Bayes classifier and ANN are capable of classifying customers according to their repsonse.

See Project!
Customer Segmentation IMG

Customer Segmentation

Python

Given transactional data for a UK-based online retail store spanning 1 year, this project implements a classic Unsupervised Machine Learning Model in order to stratify customers into three distinct groups with respect to the value they pose to the business.

See Project!
yogini IMG

yogini (DubHacks'20 Submission)

JavaScript

Your personal yoga assistant during quarantine, fully capable of monitoring form accuracy for distinct poses in real-time with seamless analysis of video input, by leveraging the PoseNet model from TensorFlow.js. Crunched this out with my amazing teammates in a 24-hour window during DubHacks 2020.

See Project!
Portfolio IMG

Personal Portfolio

HTML5/CSS3

After learning the fundamentals of web design from several resources and tinkering around with the creation of toy-websites, I finally decided to focus my efforts and build my personal portfolio-site from scratch with my bare hands(... or fingers, I suppose?), while learning a great deal along the way.

You're looking at it!

Work Experience

My experiences on my journey so far in my different positions have each prepared me in a unique way to improve my mindset and skills as an individual as well as sharpen my thought process when it comes to coordination and teamwork.

  • Huawei
  • Artificial Intelligence Researcher Intern
  • May 2021 - December 2021
  • Developed multiple features for the prototype of a data trading application, including searching, product exploration and market activity summarization (stack: Angular, Flask, PostgreSQL, Elasticsearch)
  • Currently working on Deep Learning experiments (using PyTorch) centered around a privacy-preserving data valuation scheme for ML datasets in a trading context
  • Leveraging cutting-edge cloud computing resources to train/evaluate neural networks in parallel on a large scale
  • Worked on developing a RESTful API in Golang, for a chrome extension to segment news articles based on similar textual content and political notions (Project #2, 2020)
  • Gained hands-on development experience with AWS Lambda functions for an Android app for attendance tracking, and stayed accountable by meeting weekly deadlines and participating in code reviews in an Agile environment (Project #1, 2019)
  • Built a content based image recommendation system for ground-based telescopes while working for the Herzberg Astronomy and Astrophysics (HAA) department, using Semi-supervised Deep Clustering models
  • Gained extensive experience with all components of a Machine Learning project pipeline including image preprocessing, model training/evaluation/tuning in Keras (key models: Convolutional Autoencoders, Self-Organizing Maps)
  • Designed a web-application prototype to expose the fully trained neural network as an interactive map, according to requirements laid out by senior astronomers in the department over the course of an iterative process, using Flask, D3.js, AJAX and jQuery
  • Link to publication: An astronomical image content-based recommendation system using combined deep learning models in a fully unsupervised mode
  • Worked efficiently in a team by adapting to a fast-paced, goal-driven environment. Conveyed technical jargon to and educated consumers about products in simple terms in order to make sales effectively
  • Trained new team members towards the end of job tenure in order to prepare them in the best possible way to maintain the store standards for future customers