Karanbir Singh Email & Phone Number

Toronto, Ontario, Canada

Driving Data Science projects at Magnestar, working to make Satellite Operators' lives easier, and ensuring signal connectivity for users on earth.

Greater Toronto Area, Canada

Leveraged previously unexplored data to deliver value and insights to a variety of stakeholders.Created data analytics capabilities for a $200M ecommerce business. Created data inflows in Trifacta and DatabricksDeveloped high-level business insights from granular data, using Data Science techniques in PowerBI and PythonCreated robust visualizations, actionable for a variety of stakeholders; from executives to district managers

Toronto, Ontario, Canada

Worked as an ML developer on early-stage project, fulfilling a variety of roles, from Data Scientist to Front-End Development.Optimized Classifiers and improved anomaly detection performance by 30% with SKLearn, increasing value of data extracted from raw sources to process by 40%Developed ML visualizations, authentication in React Influenced high-level management, guiding product’s future

Created Python application to run a C++ satellite image prediction program, eliminating extremely time intensive processes, to a single step data entry.Reduced active user-time by 18x, from 3+ hrs to 10 minDecreased compute power by 90% using a novel tile systemArchitected process finding images intelligently on Google Earth Engine on granular basis, improving prediction quality up to 300%, processing using Python RasterIO

Gurgaon, India

Built software with Python and SQL in an Agile team for a SaaS product providing AI+ML suggestions to pharmaceutical marketing.Designed primary ETL structure ingesting, logging, validating and outputting data to sinks with Azure Data Factory (ADF).Developed API wrappers increasing efficiency by 75% to perform common business-specific ADF tasks through code.Architected an intelligent status logging system inserting ADF pipeline statuses into SQL Server for UI display.Developed a REST API with Django to expose backend microservices for frontend interfacing.Created Dataiku API wrappers processing and transforming data with Pandas and outputting JSON for user preview.Unit tested all pieces of code before deployment

Ottawa, Canada Area

Tasked to create an application to streamline the management system for planning and executing a major military exerciseBuilt the system for an integrated event management application for ARDENT DEFENDER, a multinational military exercise training the Counter Explosive Threat Cycle.Application consisted of a self contained, future-proofed DBMS backend system and dynamic, intelligent user friendly interfaces and briefings.Liaised with explosive experts, International explosive divers and range control officers for requirements gathering for data, processes, and UI design at a conference at Canadian Forces Base BordenArchitected a complex relational database in MS AccessWrote robust scripts, which eliminated field duplication errors, automated 20% of data entry with intelligent presets and improved efficiency by over 25%.Designed the user experience to be suitable for non-technical users