Marcus Chiam Email & Phone Number

London, London, Gb

Working on the Discovery team

London, London, Gb

- Worked on the Flax team (https://github.com/google/flax), an open-source neural network library built on top of JAX that is used by users both in and outside of Google- Flax is fast, performant and is used in many of Google's next-gen AI products, as well as external projects - added new features to the Flax library to keep it up-to-date with other neural network libraries, including adding and updating neural network layers - made significant contributions to NNX (https://github.com/google/flax/tree/main/flax/experimental/nnx), a new, next-generation API for Flax that allows users to switch seamlessly between stateless and stateful modules, enabling them to take advantage of the performance gains of functional programming while coding in a more intuitive Pythonic and PyTorch-like style- fixed Flax bugs (edge cases, test failures, deprecation warnings, etc.) to ensure stability and reliability of the library

Toronto, Ontario, Ca

- Working on a data scientist team that provides technical data solutions to internal clients - Created prediction models for stock price movements, as well as coronavirus cases - Created a recommender system for suggesting which securities to trade - Created an interactive dashboard app to visualize information and location of power plants in Europe - Created scripts to routinely scrape data from websites and aggregate statistics for analyst teams

Toronto, Ontario, Ca

- Part of the RBC Capital Markets graduate rotational program- Working on a proof-of-concept project that involves applying data science techniques on trade settlement data to generate new insights and trends- Techniques include clustering, dimension reduction, autoencoders, and isolation forests- Generated a visualization tool to explore trade settlement data

- U of T’s Bioinformatics and Computational Biology research project course - Exploring novel biomarkers predictive of drug response from RNA sequencing and pharmacological profile data sets - Using machine learning techniques to investigate and develop univariate and multivariate feature associations, that are predictive of drug response

Toronto, Ontario, Ca

- U of T’s Bioinformatics and Computational Biology research project course - Creating a comprehensive database that consolidates drug, protein and gene data and predicts the relationships between each biological space - Using data imputation and machine learning techniques such as matrix factorization to make novel predictions on drug-target interaction

Toronto, Ontario, Ca

- Tasked with creating an AI-based recommender system to provide client-specific data insights for fixed-income salespeople - Managed team objectives, conducted interviews with end-users, developed pipelines for data processing and feature engineering - Created machine learning models for classifying clients and bonds, as well as predicting bonds for specific clients based on their historical data - Provisional patent filed for this product

- Taught students from preliminary level to Grade 10 RCM piano, composition and JOC curriculum

- Used autoclave to sterilize equipment and waste material- Calibrated laboratory scales and volumetric flasks

Toronto, Ontario, Ca

- Recruited ER patients for an anaphylaxis study- Required to centrifuge, pipette and isolate blood samples- Collected patient information and created data tables for the study

Toronto, Ontario, Ca

- Recruited ER patients for an anaphylaxis study- Required to centrifuge, pipette and isolate blood samples- Collected patient information and created data tables for the study

Toronto, Ontario, Ca

- Used CAD software to designed molds for the company logo and for the motion capturing glove to be 3D printed- Researched game designs to find potential strategies of incorporating the glove into pre-existing games as well as new game demos that we’d create- Conducted surveys to research potential aesthetic designs for the glove