Theia Labs

Supercharging businesses with computer vision analytics

2019 - Waterloo

• This failed for a few reasons:
  • We wanted to build around an idea that can be validated quickly, and people in retail were very hard to reach and didn't seem open to try new things.
  • This might have been one of those ideas where we built the product around the technology and we all know that's a big no-no.

Uber ATG

Perception Engineering Intern

2018 - Pittsburgh

• Projects:
  • Built a testing pipeline to test individual parts of the vision stack improving testing efficiency by 5x.
  • Working on object tracking algorithms within the Perception team
  • Solving the jumping bike problem (when card predicts that bikes will drive off the rack any time)

WATonomous

Perception Engineer

2018 - Waterloo

• Projects:
  • Developing a traffic light and sign detector and classifier
  • Developed a data aggregation pipeline to standardize numerous public object-detection datasets.

Shopify

Data Science Intern

2018 - Ottawa

• Projects:
  • Analyzed Data and built internal reports to understand consumer behaviour on the platform using SQL and Python.
  • Built a dimensional model to aggregate data on daily platform usage improving query time by 10x.
  • Developed features and build a customer maturity machine learning model to serve customized content to users and guide internal business decisions.

Waterloop

Software Lead

2017 - Waterloo

• Projects:
  • Responsible for managing over 30 software developers to create and deliver innovative design for the embedded and software in the Hyperloop the pod.
  • Wrote a research paper on the topic of Embedded-Systems in Hyperloop Pod describing design decisions in the system.
  • Created a Node.js web-socket server for bi-directional streaming of data from sensors on the pod to a group of remote devices, that allow to control the pod in real time.

Adknown

Software Engineering Intern

2017 - Guelph

• Projects:
  • Full-Stack Development of time tracking Jira plugin, used Jira APIs to effectively integrate time tracking.
  • Researching and Developing web-push notifications and a serverless client to send out batch notifications increasing notification conversion by 1.8x.
  • Creating various interactive and mobile-friendly interstitials.

TIONIX

Python Developer Intern

2016 - Kazan, Russia

• Projects:
  • Part of development team working on Russian Cloud Systems solution at TionixLabs.
  • Designed an autocomplete algorithm in Python to improve customer experience when controlling the cloud system through a command line interface.
  • Working with open source projects such as: OpenStack, Django, Sphinx, Pymongo to build a complete suite of cloud software and services such as Virtual Desktop and Compute Intances.

Software Systems in Hyperloop Pod

2018

• Overview of the Software System developed to control the pod during the high-speed launch in the in the annual SpaceX Hyperloop Pod competition. Reliance of redundant multi-nodal command execution with RAFT consensus algorithm and an on-board CAN network to enable modular design and data collection from Pod's sensors.

Drake Lyric Generation

2017

• Generating Drake Style Lyrics using LSTMs and Language Models. Over 30 forks on GitHub and 5k claps on Medium. Used by profs in Portugal to teach AI courses.

Amazon Go @RuHacks

2017

• Winners of the 2017 RUHacks Hackathon with a POC Amazon Go store built in 36 hrs. Read our story on Medium

How Much @QHacks

2017

• Solving optimization problem to find best meal for the money near the user. Tomato API integration for restaurant menu listing and Price Discrimination Algo. All packaged into a mobile app.

Tryify @Decode Hackathon

2017

• In partnership with Shopify, instagram-feed-like shopping experience Shopify merchants's products. Ruby backend with intelligent product search and Shopify checkout API integrations.

Medium API package

2017

• Medium's API to get all posts made by a user or a publisher. Over 500 downloads.

Medium API package

2017

• Medium's API to get all posts made by a user or a publisher. Over 500 downloads.

US Mass Shootings Analytics

2017

• Extensive visualization and statistical analysis of historical data on Mass Shootings in US.

Driver Face + Eye Tracking

2017

• YOLO and Haar Cascades face and eye detection and tracking CV pipeline. Open/closed eyes Deep Net classifier, critical component for safe AV testing.

ML-Utils library

2017

• Machine Learning utilities library for Python, simple image labelling web-tool

Snack Classifier

2017

• Solving problem of uneducated trips to the snack shelf in the office using IoT and Computer Vision. CNN for snack classification.

2nd. @HICC

2018

• International business case competition in Florida, HICC'19

1st. Solvers Cup

2018

• International business case competition in Budapest, Solvers Cup'18

1st. @UNICC

2018

• International business case competition in Spain, UNICC'18

Wilfrid Laurier Delegate @JDC Central

2017

• Selected to represent Laurier's Digital Strategy team at the JDCC business case competition

1st. @RuHacks

2017

• Amazon Go Implementation