S

Sanjir Salsabil

About

Detail

Machine Learning Engineer | Ex Sr. BA @ BMO | MSc in Data Science | Develop Agentic AI, Data Science & Business Analytics Solutions
Canada
Contact Sanjir regarding: 
work
Full-time jobs
Starting at USD51/hour
groups
Networking

Timeline

work
Job
school
Education
folder
Project
flag
Award
auto_stories
Publication

Résumé

Jobs verified_user 0% verified
  • Horizon Health Network
    Data Science & Business Intelligence Analyst
    Horizon Health Network
    Mar 2025 - Oct 2025 (8 months)
    In this role, I lead data-driven initiatives by designing and deploying Azure Machine Learning models and advanced analytics solutions to support strategic healthcare decisions. I collaborate cross-functionally to develop Power BI dashboards, automate reporting workflows, and migrate enterprise data to Microsoft Fabric. My work focuses on transforming complex datasets into actionable insights while ensuring data quality, governance, and self-service analytics adoption across teams.

    • Extract and transform complex datasets from enterprise databases (SNB, GNB) to uncover actionable insights and KPIs for healthcare operations.

    • Lead end-to-end data science workflows from data ingestion to model deployment, using Python,
  • Areto Labs
    Machine Learning Engineer
    Areto Labs
    Oct 2024 - Mar 2025 (6 months)
    • Implement Hugging Face Models and fine-tuned them for NSFW content moderation.

    • Evaluate LLM's Predictions to Automate tracking and outperform traditional moderation of harmful content i.e., abusive comments, Hate speech, racism, illegal streaming, Crypto scams and NSFW Bots across social media to ensure a safe online environment for everyone in every possible language.

    • I collaborate closely with the Head of Technology and other Engineers to develop automated moderation and fan engagement software powered by bespoke AI models. Our solution effectively manages social media interactions across multiple languages, creating a more inclusive environment while addressing harmful content in real time—with detection capab
  • Vosyn
    Data Scientist - Data Governance, AI & Infrastructure
    Vosyn
    Mar 2024 - Oct 2024 (8 months)
    Vosyn is a pioneering Voice Synthesis AI firm with a vision to empower individuals to create groundbreaking voice-enabled products, ultimately breaking language barriers and fostering global accessibility where innovation knows no bounds.

    Our mission is to break down language barriers, and the vision is that "Every voice and every language will be included". It's not just about translation but also about capturing emotions and feelings in conversations.

    • Collaborated with a diverse team to integrate CloudWatch, facilitating the monitoring of data drift and model performance in the production environment of a machine learning system.

    • Designed and developed Star Schema and Snowflake schema data warehouses u
  • BMO
    Senior Business Analyst
    BMO
    Oct 2023 - Current (3 years 1 month)
    Ensure client reference data is managed, accurate, complete, fit for purpose, and meets the requirements of business consumers involves data mastering, data quality management, and matching/merging of entity instances Responsible for providing data analysis and facilitating data remediation of master records Support the development and implementation of data governance policies, processes, and controls Involved in creating and maintaining automated reporting and data visualizations Conduct data analysis of client reference data from multiple internal BMO platforms Maintains business and technical metadata and maintains system and process documentation Provide subject matter expertise in data analytics and the client data domain to support d
  • BMO
    Senior Business Analyst (Capital Market - Master Data Management)
    BMO
    Sep 2023 - Mar 2024 (7 months)
    This role often requires me to transition between analytical tasks to leadership, act as a point of contact between Senior management, technical leads and business stakeholders while showcasing my Business enabling expertise

    Lead cross-functional CMMDM and Data Governance teams aligned with business, technical requirements and gained professional expertise in technology transformations and digitalization from inception to production

    • Conduct data analysis of client reference data (CRD) from multiple internal BMO platforms
    • Maintains business and technical metadata, maintains Data backup and process documentation
    • Developed Spark jobs in Azure Databricks Notebook for data cleaning and standardization, impr
  • Minerva Analytics Inc
    Data Scientist
    Minerva Analytics Inc
    Jun 2022 - Sep 2023 (1 year 4 months)
    Aerial Detection | Computer Vision | Deep Learning | Diffusion Model

    Targeted Consumers: Military intelligence, war fighters, and decision-makers

    Objective To develop an AI-driven satellite surveillance system capable of detecting and tracking hidden military assets, enhancing situational awareness for military intelligence and decision-makers

    o We Identify unusual activities in satellite imagery without relying on predefined patterns.
    o Tracks and categorizes objects within video frames to distinguish between threats and normal activities.
    o Detects concealed military assets by analyzing track markings and determining their path and direction.
    o Utilizes extensive satellite imagery for model
  • University of Alberta
    Data Science Researcher
    University of Alberta
    Sep 2021 - Jan 2023 (1 year 5 months)
    I collaborated in a Data-Driven Healthcare Affordability Project for Cost Prediction and Financial Planning. Aim to raise awareness about health insurance claims with Group Retirement policies by minimizing costs and maximizing benefits through informed decision-making

    •Designed Star Schema and Snowflake data warehouse structures to optimize data ingestion by 30% and reduce storage costs by 10% on records of hospital patients
    •Evaluated metrics and contrasted the performance outcomes of the most widely used machine learning algorithms
    •Harnessed advanced machine learning techniques to enhance the precision of health insurance cost predictions and reduced pricing errors by 22%

    Most notably we interpreted the
  • Minerva Analytics Inc
    Machine Learning Engineer
    Minerva Analytics Inc
    Jul 2021 - Jun 2022 (1 year)
    Collaborated with the Research and Development team to develop ML algorithms (regression, clustering, decision trees) for efficient data analysis, deployment and support end-to-end solution architecture

    Real-Time Data Pipeline Architect: Designed and deployed scalable data pipelines using Apache Spark, Kafka, and Airflow on MS Azure ML Studio to handle large volumes of real-time streaming data for customer profiling and improved geolocation features. This resulted in a 30% increase in customer engagement and a 20% rise in repeat purchases

    Large-Scale Data Management and Scalability: Collaborated with stakeholders to manage diverse data formats and sources, implemented a scalable data infrastructure ensured efficient ha
  • Minerva Analytics Inc
    Data Engineer
    Minerva Analytics Inc
    Jul 2020 - Jul 2021 (1 year 1 month)
    Armed with deep cloud proficiency and Machine Learning expertise, played a pivotal role in demonstrating the transformative potential of AWS, and Azure services for our clients' reference data challenges

    • Led a team of 15+ data analysts, data scientists, and BAs, driving AI/ML projects and showcasing analytics expertise in pre-sales engagements
    • Prepared data quality checks, including data validation for completeness, duplicate removal, and handling missing values
    • Refined risk prediction models using PySpark and Azure Machine Learning on a 5 million consumer database
    • Investigated critical factors influencing cost fluctuations, and applied machine learning models achieving an 18% increase in cost prediction
  • Dotlines
    Data Analyst
    Dotlines
    Jul 2019 - Jun 2020 (1 year)
    Successfully implemented the company's own digital platform Audra (https://audra.io/) in response to market trends and competitors' advancements. Integrated the system payment gateways (using APIs) and intuitive User Interface. Provided ongoing maintenance and support for the system, including logging, and monitoring

    Coordinate with team members, work under a time-sensitive, pressured and fast-paced dynamic environment and demonstrate my expertise in transitioning older systems to modern platforms while ensuring seamless functionality

    •Overcome the challenges of navigating legacy applications, outdated infrastructure and compatib
  • KPMG
    Data Analyst
    KPMG
    Feb 2019 - Jul 2019 (6 months)
    Project Title: Retail Inventory Management Optimization through Customer Segmentation and Demand Forecasting

    • Simplified real-time data visualizations and daily sales reports using SharePoint and Power Automate Flow to improve data accessibility and reduce report generation time by 40%
    • Evaluated customer data using Python and PySpark to identify distinct customer segments, uncovered 3 unique buying patterns and habits
    • Segmentation qualifies targeted marketing campaigns, leading to a 10% increase in customer acquisition
    • Employed time series forecasting models to predict future demand with an estimated 95% accuracy
    • Reduced Inventory Holding Costs by 20% and generated a comprehensive global retail mark
  • Accenture
    Business Intelligence Analyst
    Accenture
    May 2018 - Aug 2018 (4 months)
    Project Title: Enhance Digital Marketing a Multi-Faceted Approach to Content Trends

    • Cleaned, modelled, and analyzed 5 datasets to uncover insights into High performance direct to consumer marketing content trends
    • Achieved 15% increase in content engagement (metrics include clicks, shares, comments, and time spent on content)
    • Redesigned Power BI dashboards to visualize customer feedback and user data to facilitate a 20% increase in content-focused integrated marketing strategy
    • Data insights led to create more video content, which resulted in a 20% increase in user engagement and click-through rates
    • Key data insights from the SharePoint report and presentation earned a "Highly Informative" rating fro
  • PwC
    Power BI Developer
    PwC
    Feb 2018 - May 2018 (4 months)
    Project Title: An Automated HR Data Analytics Platform to monitor Workflow management and Onboarding Resources

    • Co-ordinated employee onboarding processes and a 10% reduction in time-to-hire

    • Tackled data analysis challenges on three projects and overcame data quality issues like incompleteness, inconsistency, typos, outliers, duplicates, and analytical challenges such as identifying the right metrics and data bias
Education verified_user 0% verified
  • University of Alberta
    Master of Data science & Big Data Analytics, Mathematics and Computer Science
    University of Alberta
    Sep 2021 - Jan 2023 (1 year 5 months)
  • IDP Education
    International English Language Testing System, English Academic & Professional
    IDP Education
    Feb 2021 - Feb 2023 (2 years 1 month)
    Advanced User. Competent Level C1
  • IDP Education
    International English Language Testing System, English Academic & Professional
    IDP Education
    Jan 2021 - Dec 2023 (3 years)
    Advanced User. Competent Level C1
  • International Islamic University Malaysia IIUM
    Bachelor of Applied Science - BASc, Computer Science
    International Islamic University Malaysia IIUM
    Jan 2015 - Dec 2019 (5 years)
    Assistant Head of International Affairs at ICT Student Society (IIUM_ICTSS)
    Project Co-ordinator at IIUM Incubator Advanced Tech Sdn Bhd
  • International Islamic University Malaysia
    Bachelor of Applied Science - BASc, Computer Science
    International Islamic University Malaysia
    Jan 2015 - Dec 2019 (5 years)
    Assistant Head of International Affairs at ICT Student Society (IIUM_ICTSS) Project Co-ordinator at IIUM Incubator Advanced Tech Sdn Bhd
  • S
    Higher Secondary School Certificates, Science
    S O S Herman Gmeiner School College Dhaka
    Jan 2012 - Dec 2014 (3 years)
  • British Council
    Pre-intermediate English Spoken Course, English Language and Literature, General
    British Council
    Jan 2012 - Dec 2012 (1 year)
  • S
    Secondary School Certificates, Science
    S O S Herman Gmeiner School College Dhaka
    Jan 2002 - Dec 2012 (11 years)
Projects (professional or personal) verified_user 0% verified
  • R
    Real-Time Healthcare Analytics Strategic Performance Dashboard
    Mar 2025 - Sep 2025 (7 months)
    Built enterprise-level Power BI dashboards featuring 17 interactive reports, visualize real-time metrics across 5 strategic pillars and tracking 20+ KPIs across community care, hospital excellence, workforce management, research innovation, and patient safety.

    This data-driven solution empowered executive decision-making and contributed to measurable improvements in surgical wait times, operational and patient care outcomes for a network serving multiple hospitals.
  • A
    Azure Open AI | Machine Learning Predictive Analytics on Healthcare Performance Metrics | Responsible AI
    Mar 2025 - Oct 2025 (8 months)
    Data-Driven Surgical Capacity Optimization: Multi-Year Backlog Elimination Strategy

    Developed comprehensive predictive modeling framework using Azure Machine Learning. Analyzed 12,531+ surgical waitlist cases across 4 zones, multiple specialties, Facilities and patient demographics to address critical capacity crisis.

    Forecasted 5-years trajectories showing potential growth to 23,000 patients without intervention, and modeled strategic scenarios demonstrating that 10,000-15,000 additional annual procedures ( with 20-35% capacity increase) sustained over 2-5 years could systematically eliminate backlogs.

    Delivered evidence-based, targeted recommendations, prioritizing high-demand specialties like Orthopedic
  • A
    AI-powered fine-tuned ViT models to combat 5,000% surge in Ticket Scam & Betting Model
    Jan 2025 - Mar 2025 (3 months)
  • A
    AI-Powered NSFW Image Detection System for Social Media Content Moderation
    Oct 2024 - Dec 2024 (3 months)
    Developed an automated NSFW detection service at Areto Labs to identify inappropriate profile images used by bots targeting young athletes on social media.

    Integrate computer vision models with an existing microservices architecture.
    Optimize detection accuracy while minimizing false positives.
    Real-time processing of scraped Instagram profile images.
    Manage embedded content and variations in image quality.
    Fine-tuned Base NSFW models from Hugging Face (FalconsAI) PyTorch and Google Vision API.

    Key Accomplishments:
    o Achieved 99% overall accuracy using fine-tuned Hugging Face ViT models.
    o Optimized confidence threshold (0.4 vs 0.5 default), enhancing F1-score to 0.97.
    o Reduced the f
  • E
    Examining the effect of contact rate and stability of the SIS Epidemic model for N. Gonorrhea and SIT model for HIV tran
    Jan 2022 - May 2022 (5 months)
    • I conducted a thorough review of the existing literature to identify key research questions, gaps in existing knowledge, future works, and potential areas for my research and experiment.
    • Analyzed the data using RStudio, and STATA and developed mathematical models to simulate disease transmission, conducted sensitivity analyses to explore the problem and robustness of my research findings, and visualized the results using graphs in MATLAB.
    • Interpreted a clear description of my research question, methods, results, and conclusions, including a discussion of the implications of my findings for public health policy and practice.
  • S
    Statistical Methods for Applied Research II (STAT 537)
    Sep 2021 - Jan 2022 (5 months)
    Our in-depth research is based on “A new method for Espresso Coffee brewing: Caffe Firenze,” published in the Journal of Agricultural Engineering.

    The data we used is simulated data shares the same mean and standard deviation. The dependent variables are temperature and pressure. The dependent variable is the ratio of Crema to liquid in percentage.

    Our null hypothesis is that the temperature and pressure have no effect on the crema ratio.
    While our alternative hypothesis is that either – or both – of pressure and temperature influence the ratio of crema.
  • M
    Modelling the Movement of Salmonella in Gastrointestinal Mucus
    Sep 2021 - Dec 2021 (4 months)
    In our project, the movement of Salmonella bacteria in fresh, undiluted gastrointestinal mucus frommice clearly examined. Accomplished by examining 993 readings of different bacteria particle movements from the previous study done bySchroeder et al.

    These literature review are from 27 different videos where the movement of individual Salmonella bacterium were tracked, with their x and y coordinates measured in pixel coordinates of dx= 0.156 μm and the time measured in frames of dt= 0.0667seconds.
  • A
    Application of Big Data Analytics in Predicting on-Time Flight Performance.
    Feb 2019 - Sep 2019 (8 months)
    Airline delays are existent in all parts of the world. This research aims to identify whether a flight will
    experience delay depending on various parameters including the month, airline, day of the week, flight number, the origin of the airport, a destination of the airport, and scheduled departure of a flight. Some methods exist in the literature for predicting on-time flight performance. From the peer review of the existing research paper of this area, implementing the gradient boosted decision tree algorithm has yielded prediction accuracy results. This research proposed the implementation of A Two-Class Boosted Decision Tree to achieve significant accuracy in prediction compared to existing algorithms.
    Additionally, Users can
  • I
    Image To Code in Machine Learning
    Sep 2018 - Dec 2018 (4 months)
    The "Image to Code" project represents a significant advancement in the field of machine learning, enabling the automatic transcription of handwritten code into compilable digital code. This technology greatly benefit educational institutions, making it easier for instructors to evaluate handwritten code submissions accurately and efficiently. By leveraging Azure Machine Learning, deep learning, and computer vision techniques, this project aims to bridge the gap between handwritten and digital text in the context of programming languages.


    Model Selection, Training & Evaluation:
    We Use a deep learning-based OCR model (e.g., CRNN, Tesseract with LSTM) to recognize and transcribe handwritten text into digital text.
  • D
    Data Science Project on Health Insurance
    Sep 2018 - Dec 2018 (4 months)
    One of the major concerns is our health, and this is where health insurance comes into the picture. Although many people don’t like the idea of having to pay for insurance, but it can be a life saver. My data science project aims to help both insurance companies as well as normal citizens. Insurance companies would find it easier to figure out how much they should charge customers that possesses certain criteria such as Smoking Habits, BMI, Age, No. of Children etc. To predict the outcomes of my research I train a model using a machine learning algorithm called two class boosted decision tree.
Awards verified_user 0% verified
  • K
    KPMG Cyber Security Challenge 2018
    Oct 2018
  • D
    Deans List
    Sep 2018
  • C
    CAT HACK Prototype Designing Contest
    Jan 2018
    First Runner Up
  • A
    ACM ICPC
    Jan 2017
    Programming Contest Participation Award
  • A
    Academic Excellence Award
    Jan 2017
  • D
    Deans List
    Jan 2016
  • D
    Deans List
    Sep 2015
Publications verified_user 0% verified
  • A
    Optimized Bandwidth Recycling in IEEE 802.16 Networks
    Adamu Abu Bakar Ibrahim Dr Sanjir Inam Salsabil
    Jan 2019