Summary
Overview
Work History
Education
Skills
Accomplishments
Timeline
Generic

Satyake Bakshi

Toronto

Summary

2x Azure Associate Data Scientist


Skilled Data Science and analytics professional well-versed in utilizing Tensorflow/PyTorch, ML Stack to develop advanced machine learning models. Certified in the use of hyperscalar solutions like Google Cloud Platform (Certified) and Azure Machine Learning Platform (Certified), to deliver impactful solutions. Eager to contribute my expertise to dynamic projects and leverage data-driven insights for success.


Status in Canada: Permanent Resident



Overview

4
4
years of professional experience

Work History

Data Science Developer

Bell Media | BCE Inc.
05.2023 - 02.2024
  • Frontend Dynamic Microstrategy dashboard development for Bell's CRAVE platform.
  • Working with various sub-vendors: Apple, Roku, and CSG in developing end to end ETL /ELT development and scheduling/deployments in GCP using BigQuery, Cloud Functions, Cloud Composer, VertexAI service .etc (GCP Stack) as part of Bellmedia's critical MDM (Media Data Modernization) program.
  • Designed end to end ETL for business reporting using, Bigquery, Dataproc, Cloud functions .etc
  • Worked with offshore QA teams for both validations of backend and front-end loads.

    Tools used: Google Big Query, Google Cloud Functions, Vertex AI service, Cloud Composer, Cloud Data Fusion, DataProc Clusters, Microstrategy Reporting

Sr. Associate, Credit Risk Models (Full-Time)

FEM |Deloitte, BAE
10.2022 - 03.2023
  • Worked with the Credit Risk Modelling team under Financial Advisory Line of Business
  • Involved in the Review and Validation of multiple credit risk models, including Probability of Default(PD), Loss Given Default(LGD) etc.
  • Worked with external partners in model validation, validation of methodologies, performance metrics, monitoring regiment with Bank of the West, BNP Paribas, BMO.


Quantitative Analyst-Data Scientist (Full-Time)

Model Risk Management | TD
08.2021 - 08.2022
  • Performing Model validation of various High-Risk Fraud Models >$ 20 million.
  • General Model assessment utilizing statistical metrics: PSI AUC, AUPR, and other statistical tests.
  • Evaluation of Fraud, CyberSecurity models.
  • Development and Evaluation of Deep Learning based Fraud Detection Models for financial loss minimization.
  • Annual Model Reviews of various internal/vendor models.
  • Performing Performance monitoring of various vendor models developed by VISA, FICO.etc using Model Validation (MV)-approved metrics.
  • Escalation of high-risk models, determined by Model Risk Management Framework(s).
  • Tools used: Python, Pandas, JupyterLab, Scikit-learn, etc.


Graduate Researcher

Carleton University
09.2019 - 09.2020
  • Investigation of Protonets and few-shot architectures in predicting the response of biomimetic materials for smart camouflage (DNDC/NSERC).
  • Development of a Novel parameter-efficient CNN for learning feature representations (DNDC/NSERC).
  • Development of a Fused Inception-Densenet architecture for detecting Human Falls (DNDC/NSERC).
  • Development of LSTM/CNNs/Few-Shot models using M-mode and B-mode images of ultrasound obtained from the Carleton imaging lab.
  • Realtime Measurement during Systole and Diastole of Arterial Diameter, using TKEO for detection and treatment of Atherosclerosis.
  • Tools used: Python, TensorFlow, PyTorch, Scikit-learn, Azure ML SDK, Azure Compute Clusters.

Education

Master of Science - MAScBiomedical Engineering

Carleton University
Ottawa, ON
05.2021

B Tech - Biomedical Engineering

VIT University
05.2018

Skills

  • Deep Learning
  • Machine Learning
  • Natural Language Processing
  • Python, TensorFlow, PyTorch stack (ML/DL Stack)
  • SQL
  • Azure ML (Studio SDK)

Accomplishments

    • S. Bakshi, S.Rajan, “Few-shot fall detection using shallow Siamese network,” in 2021 IEEE International Symposium on Medical Measurements and Applications (MeMeA) (MeMeA 2021), Neuchatel,Switzerland, June 2021.


    • S. Bakshi and S. Rajan, "Fall Event Detection System using Inception-Densenet Inspired Sparse Siamese Network," in IEEE Sensors Letters, doi: 10.1109/LSENS.2021.3089619.


    • Bakshi, Satyake. Investigation of Few-Shot Learning for Fall Detection. Diss. Carleton University, 2021.


    • Bakshi S. Attention Vision Transformers for Human Fall Detection. Research Square; 2022. DOI: 10.21203/rs.3.rs-1614908/v1.

Timeline

Data Science Developer

Bell Media | BCE Inc.
05.2023 - 02.2024

Sr. Associate, Credit Risk Models (Full-Time)

FEM |Deloitte, BAE
10.2022 - 03.2023

Quantitative Analyst-Data Scientist (Full-Time)

Model Risk Management | TD
08.2021 - 08.2022

Graduate Researcher

Carleton University
09.2019 - 09.2020

Master of Science - MAScBiomedical Engineering

Carleton University

B Tech - Biomedical Engineering

VIT University

Similar Profiles

CREATE PROFILE
Satyake Bakshi