Senior Data Scientist, Platform Implementation
Chain-Sys Corp.
Total years of experience :5 years, 11 Months
Developed and executed a comprehensive multi-language expansion strategy, resulting in a 200% increase in user base within 6 months.
Successfully completed multiple state-of-the-art Large Language Models (LLMs) Prompt Engineering POCs resulting in the productionization of cutting-edge Generative AI solutions that improved customer satisfaction by 40%, while reducing core model maintenance needs by 70%.
Designed and implemented retraining pipelines for the BERT models used for Adversity categorization.
Transformed data science prototypes into production-level solutions on the Azure suite, converting machine learning models into APIs and building a model-as-a-service platform, resulting in a 300% increase in model accessibility and a 40% reduction in time-to-market.
Led labeling initiatives by overseeing teams of compliance researchers, and data operations personnel, resulting in an increase of labeled data by 600% and enabling the development of more accurate models.
Led the development of data drift and model monitoring mechanics using tools like mlflow and Azure Functions, leading to a 20% decrease in data drift incidents and a 15% increase in model reliability.
Implemented a robust monitoring pipeline to identify and address data, model, label, and prediction drifts, reducing errors by 30% and improving overall model accuracy.
Led the successful roll out of 5 Government of Canada funded digital health solutions in 2 years with Big Data capabilities, consumer initiatives, and analytics framework.
Oversaw the activities of a team of Data Scientists and Full Stack Developers, ensuring alignment with lab objectives and achieving a 30% increase in project efficiency.
Utilized Python programming language and statistical models to analyze geographic information systems (GIS) data from Google Maps API, leading to actionable insights that improved the accuracy of patient location tracking by 25%.
Developed and implemented a scalable data infrastructure using Microsoft Azure, Google Cloud Platform, and Amazon Web Services to support the analysis of large datasets, resulting in a 30% increase in efficiency of data processing.
Led the product development team in creating a new mobile application utilizing Flutter framework for seamless data collection and analysis, resulting in a 40% reduction in time spent on manual data entry.
Implemented the backend for both platforms using PHP, and SQL, following best practices to write efficient insertion, update, and retrieval queries.
Led 2 teams to develop and implement data-oriented AI-backed public health service apps to address COVID-19 readiness and mental health.
Researched and proposed platform frameworks using deep learning models like ANN and RNN towards designing a federated ML model that adapts user re-engagement based on user attributes and app usage patterns.
Implemented the backend for both platforms using PHP, and SQL while following best practices to write efficient insertion, update, and retrieval queries.
Researched and deployed AWS services to optimise data flow, with special concentration on data privacy and security needs through efficient traffic monitoring and control using Security Groups.
Designed data pipelines for both platforms in Python using data cleaning, processing, and visualisation libraries like NumPy, Pandas, and Matplotlib.
Increased platform UX/UI performance approval from 30% to 80% and thereby improved user re-engagement, by designing and developing improved user interaction flows and creating custom widgets using Flutter.
Communicated and coordinated the data needs and privacy restrictions for each platform between design and development teams by designing workflows and role distribution.
Established end-to-end data infrastructure, from data collection to storage, optimizing the ETL process and reducing data processing time by 10%.
Led the design and implementation of a scalable data warehousing schema, resulting in a 15% improvement in query performance and overall data accessibility.
Implemented automated pipelines for data cleaning, preprocessing, and feature engineering, reducing manual effort by 50% and accelerating model development timelines.
Developed a real-time dashboard that automatically generates key business insights, improving decision-making efficiency by providing instant access to critical information.
Introduced machine learning models for predictive analytics, leading to a 5% reduction in operational costs through
optimized resource allocation.
Proficient in designing and implementing data solutions using technologies such as SQL, Python, and
Spark.Implemented cloud-based data storage solutions, leveraging platforms like AWS and Azure to enhance
scalability and accessibility.
Implemented robust data governance policies, ensuring data integrity, security, and compliance with industry
regulations.
SFU Multimedia Lab:
http://multimedia.fas.sfu.ca/
Developed a Python library using Cython to facilitate Deep Feature Tensor Compression in a collaborative AI framework.
Modified 3 image compression codecs, i.e., JPEG, PNG and JPEG-XR to compress feature tensors.
Performed model performance study to analyse and tune out any undesirable effects of compression on model performance using TensorFlow, and Keras.
Cleaned, processed, and visualised image data using OpenCV to sample training, and test datasets to study the effect of noise on model performance.
Defended the project thesis successfully, in front of MEng. Project Completion panel at Simon Fraser University.
Conducted in-depth financial data analysis to support decision-making processes, identifying key trends and insights
crucial for budget planning and optimization.
Assisted in tracking advertising campaign budgets and contributed to the development of accurate financial forecasts,
ensuring alignment with strategic objectives.
Conducted variance analysis on advertising expenses, identifying discrepancies and proposing corrective actions to
maintain budgetary control and cost-effectiveness.
Utilized Excel and other financial modeling tools to create detailed financial models for advertising campaigns,
aiding in scenario analysis and providing insights into potential cost-saving opportunities.