Experience

 
 
 
 
 
Feb 2019 – Present
Toronto, Ontario

Decision Lead Developer, Machine Learning

Kwikee Labs


  • Product Image classification using TensorFlow 2 (1M datasets into 40 categories)
  • Bounding box identification using TensorFlow 2 (AutoML replacement work)
  • Product classifier (custom) implemented in the batch processing and distribution channels
  • Global printer classification and identify anomalies, Quick Search with NLTK.
  • ETL work on Global products and classification.
  • Distribution Engine Architectural redesign including customized data engineering and visualization models.
 
 
 
 
 
June 2018 – Present (Part time)
Toronto, Ontario

ML Research Lead

Tact Labs


  • Technologies Used: Python, SciPy, Pandas, Flask, ORM, Amazon AWS, REST
  • Similar developer prediction based on developer learning, skills (Similar to Shazam algorithm)
  • Developer skills authenticator (KNN & Customized Bayes algorithms) and working on the research paper.
  • Error Classifier - Based on NB algorithm to classify the error to be in Python or Java or other.
  • Developer Simple search with NLP.
 
 
 
 
 
Feb 2015 – Jan 2019 (4 years)
Minneapolis, MN

Application Architect

Smile Gate West


  • Gamer, Weapon classification and buyer prediction prototype with 87% accuracy rate.
    • Social health check analyzer provided using ML Algorithms & NLTK libraries (Vader).
  • Buyer prediction introduced by analyzing 7-year data collection.
  • New Core APP Architecture introduced to support independent Layers.
  • Reduced Third party dependencies and Cost up to 30%(Iovation) on Security and Hacker Identifier
  • Implemented Microservices as a research and POC work.
  • APP Business policies segregated on Layer level to avoid coupling issues.
  • Buyer classification and prediction introduced with KNN support.
  • Provided various ML prototypes to check buyers’ health and activities.
 
 
 
 
 
Oct 2017 – Oct 2018
Online

Co-Author - Building RESTful Web Services with Spring 5

Packt Publications


  • Technologies Used: Spring Boot, Spring REST, Spring AOP
  • Simplified all Spring components with the updated Spring Boot and Maven version.
  • Targeted for Intermediate Spring REST audience for migration and tech upgrading purpose.

Projects

*
All Machine Learning MLOps (MaaMs)

Product Classification and Validation

Predicting 12,000 product classification and validation

Customer Transcation Prediction

Identifying which customers will make a specific transaction in the future

Project 2

Project content 2

Project 3

Project 3 content

Project 4

Project 4 Content