CV

Machine Learning

85%

Embedded Systems

75%

DevOps

70%

Skills
I consider myself a philomath; a person who enjoys learning new facts and acquiring new knowledge. My self-motivated nature and curiousity make me want to understand many things.

Hardwork

100%

Education
  1. MASc, Machine Learning and Artificial Intelligence
    University of Guelph Supervisor: Dr. Graham W. Taylor
  2. BEng, Engineering Systems and Computing
    University of Guelph

Experience
As a guy with a "systems-thinking" mindset, I embrace different paths to pursue diverse interests and open more doors. This hollisitic approach lead me to working in various industries.
  1. Data Scientist
    I worked with a small team of Data Scientists on a Chatbot project using Deep Learning techniques. I spent a lot of my time preparing conversational corpuses as training data and contributed to the generative conversational model. Used: TensorFlow, Anaconda, NLP, Django, Docker, AWS.
  2. Hardware & Systems Dev
    I worked mostly independently on a power and clock calibration library for ON Semiconductor's latest multi-protocol Bluetooth 5.0 SoC (RSL10). I also performed hardware and firmware verification for the chip's BLE:GAP pairing and bonding. Used: C, ARM Assembly, Eclipse, J-Link, nRF Studio.
  3. Software Engineer
    Developed a Bit Estimator algorithm with 91% accuracy, as well as an Artifact Reducing Filter for the HEVC encoder software library. I also, researched and implemented a novel Capped Variable Bitrate (CVBR) algorithm for Real-Time H.264 video encoders and transcoders. Used: Python, C, Visual Studio, SVN.
  4. Android App Dev
    I worked with the Architecture and Technology team to implement security libraries for secure D2D communication in the core of an Android IoT application. Also, explored Smart Home Automation APIs from Nest, Honeywell and Apple to be used for the app. Used: C, Android NDK, GDB.

Projects

  1. Seq2Seq Regression
    Novel Deep Learning API in TensorFlow for regression.
    Worked with Thorsteinn H. Jonsson on a sequence-to-sequence (seq2seq) code in TensorFlow that can be used as a tool for non-linear regression analysis or performing a wide range of automatic feature extraction tasks outside of NLP.
  2. Weather Forecaster using ANNs
    Make weather predictions all year round using this API.
    It's a lot harder to predict the weather these days according to meteorologists. Implemented a model using Artificial Neural Networks (ANNs) in MATLAB for predicting the weather for a city with high tourist traffic at 75% accuracy.
  3. Solar-Powered Roadside Monitoring System
    Intelligent vision system to help prevent construction worker deaths.
    As the amount of motorized vehicles increases, the safety of roadside construction workers has been negatively affected. This system monitors the presence of road-side construction workers using computer vision and Beacon technology, then adjusts the speed-limit sign in real-time.