Efe Tascioglu

Profile

Engineering Science graduate majoring in Machine Intelligence, with experience writing scalable code while applying and developing algorithms for computer vision, reinforcement learning, and natural language processing.

Experience

Amazon

Toronto, ON

Software Development Engineer

May 2025 - Aug 2025

  • Optimized startup latency on low-memory Alexa devices, reducing initialization time from 13s to 1s by redesigning system ownership
  • Designed and developed memory-profiling tools for FireOS, including interactive flame graphs for heap allocation analysis
  • Detected and patched memory inefficiencies in C++, saving more than 2x the memory desired in the project proposal
  • C++
  • FireOS
  • Memory Optimization
  • Performance
  • Alexa

Huawei

Toronto, ON

Machine Learning Researcher

Sep 2024 - Apr 2025

  • Designed and distilled large language models (LLMs) to implement Speculative Decoding, achieving a lossless 3x performance increase
  • Designed and developed LLM quantization strategies using EfficientQAT, outlier removal, mixed-precision quantization, double-quant, and vector quant to achieve 4x speed and memory improvement with 8% rise in perplexity
  • Python
  • PyTorch
  • LLMs
  • NLP
  • Quantization
  • Model Optimization

Amazon

Toronto, ON

Software Development Engineer

May 2024 - Aug 2024

  • Designed and developed an improved versioning architecture for Alexa using C++, directly reducing compatibility errors for over a dozen teams
  • Developed a new low-maintenance notification and API test suite for inter-package communications preventing future downtime
  • Contributed to the Alexa Hybrid Engine to reduce device failures, improve unit test coverage, and simplify existing codebase
  • C++
  • System Design
  • Alexa
  • Testing
  • API Design

NerveX Neurotechnologies

Toronto, ON

Artificial Intelligence Developer

May 2023 - Aug 2023

  • Performed feature engineering and unsupervised anomaly detection to detect seizures in epileptic patients for improved treatment
  • Designed algorithms to denoise biosignals and extract vital signals with 50% higher accuracy using noise detection and removal methods
  • Developed and deployed microservices & algorithms at scale using AWS and Docker, and optimized existing algorithms to perform 3x faster
  • Python
  • Signal Processing
  • ML
  • AWS
  • Docker
  • Biosignals

Tunescape

Remote

Co-Founder, Full Stack Developer

May 2022 - Dec 2022

  • Founded a startup helping individuals discover new music on both web and mobile through the analysis of technical characteristics and AI
  • Developed the backend for the application using Flask with Python, and frontend for the supported platforms using Flutter
  • Python
  • Flask
  • Flutter
  • Full Stack
  • Startup
  • Mobile

Education

BASc in Engineering Science - Machine Intelligence

University of Toronto

Sep 2021 - Apr 2026

Cumulative GPA 3.84 / 4.0 | Dean's List (6x)

Projects

Master's Thesis - SUDEP Diagnosis

2025

  • Sudden Unexpected Death in Epilepsy (SUDEP) diagnosis from ECG data, using cross-frequency coupling and machine learning
  • ML
  • Signal Processing
  • Healthcare
  • Research

UofT Machine Intelligence Student Team (UTMIST)

Sep 2022 - Apr 2024

  • Technical Project Manager, Director, Reinforcement Learning Developer
  • Evolved and trained creatures with custom morphologies and neural structures using reinforcement learning in Unity
  • Lead the design, development, and management of a new multiagent collaboration model
  • Reinforcement Learning
  • Unity
  • C#
  • Leadership

Hack the North 2023 - Iridis

Sep 2023

  • Developed a physical heads-up display helmet that records attendee faces and information to display when you encounter them
  • Used computer vision to scan and recognize faces, created a realtime database to store users for further recollection
  • Designed a concurrent backend API and deployed code to create a fast, scalable system enabling rapid prototyping
  • Computer Vision
  • OpenCV
  • Python
  • Real-time
  • Git

Hack the North 2022 - Happy Hours

Sep 2022

  • Designed and built an online tool to alleviate difficulties in organizing social events by automating activity recommendations and scheduling
  • Transformed favorite activities into locations and slotted them into a dynamic timetable using natural language processing
  • Integrated with Cohere for NLP and Google Maps API to create a roadmap of scheduled events
  • Deployed backend using Docker on Heroku creating a fast, scalable system
  • Python
  • NLP
  • Flask
  • Docker
  • Heroku
  • API

UTM AI Competition 2022 - Finalist

Jan 2022 - Apr 2022

  • Achieved 80% accuracy identifying corporate financial risk using random forest classifier on real dataset from CIBC
  • Performed data cleaning and feature engineering to achieve a more accurate model with reduced over-fitting
  • ML
  • Python
  • Pandas
  • Random Forest

Touch Guitar

Jan 2022 - Mar 2022

  • Designed a touch-screen add-on to guitar to lessen pressure required to press frets for arthritis patients
  • Accomplished 97% reduction in force required while maintaining 89%-94% of original sound quality
  • CAD
  • Prototyping
  • Hardware
  • Design

Daisy Intelligence Hackathon 2022

Jan 2022

  • Trained a regression model identifying major indicators of heart disease from among 70 causes
  • Discovered parameters critical to model confidence that were not part of the 14 variables commonly used in literature
  • ML
  • Regression
  • Python
  • Data Science

UTEK Competition 2022 - Finalist

Jan 2022

  • Designed and created a prototype robot for trash collection capable of capturing various forms of garbage
  • Performed in top 5 among 24 teams in terms of speed, capacity, and efficiency of trash collection
  • Robotics
  • Design
  • Engineering

Text to Braille - Canada Science Fair Silver

Sep 2018 - Jun 2021

  • Received Gold Medal in Toronto Science Fair and Silver Medal in Canada Wide Science Fair
  • Featured in Make Magazine as a Cutting Edge Contribution for building a reading aid for the visually impaired
  • Created a device to capture text images and convert to braille and audio in real-time using Raspberry Pi, Arduino, and OCR
  • Python
  • Raspberry Pi
  • Arduino
  • Computer Vision
  • Hardware
photo of me

Skills

    • Programming Languages
    • Python
    • C++
    • C
    • C#
    • JavaScript
    • RISC-V
    • Flutter
      ML/AI Frameworks
    • PyTorch
    • HuggingFace
    • OpenCV
    • TensorFlow
      Tools & Platforms
    • Git
    • Docker
    • Linux
    • AWS
    • NodeJS
      Frameworks & Libraries
    • Flask
    • Django
    • React
    • Unity

Courses

  • (ROB501) Computer Vision for Robotics
  • (MIE451) Decision Support Systems
  • (CSC413) Deep Learning
  • (ECE353) Operating Systems
  • (ECE454) Software Optimization
  • (ROB311) Reinforcement Learning
  • (ECE358) Algorithms and Data Structures

Awards

  • Dean's List (2024)
    Awarded 6 times
  • Finalist - CIBC AI Competition (2022)
  • Finalist - UTEK Competition (2022)
  • Amateur Rocket Certification (2022)
  • Silver Medal - Canada Science Fair (2021)
  • Featured on Make Magazine (2021)

Interests

  • Machine Learning
  • Computer Vision
  • Reinforcement Learning
  • System Optimization
  • Natural Language Processing
  • Open Source