Anna(Xiaoying) Zhao
Concord,ON
Summary
Enthusiastic data scientist/machine learning engineer with 5+ years of experience in leading data science
teams across multiple divisions, developing machine learning models, designing end-to-end pipelines, and
transforming data science prototypes to production-grade solutions.
Highly accomplished in solving critical problems for Marketing, Finance, Planning, Risk Management teams in Retail and Insurance industries, influencing decision-makers and leadership, reducing operational costs and driving profitability. Eager to bring expertise to growing organization in challenging new role.
Overview
15
15
years of professional experience
Work History
Senior Data Scientist /Machine Learning Engineer
Financial Services Regulatory Authority Of Ontario
09.2023 - Current
- Design and develop the enterprise-level LLMs with Langchain and Azure OpenAI for Credit Union Sector, customize the functionalities with ensemble
- Enhance and upgrade end-to-end AI solutions for the Decision Support Portal to help Mortgage and Insurance Sectors to conduct massive background checks and auto-review documents for licensing with the most advanced NLP algorithms and LLMs.
- Collaborate with different Financial Sectors to design the roadmaps and methodologies for the upcoming NLP projects with LLMs.
Data Scientist /Machine Learning Engineer
HBC ( Hudson's Bay )
06.2021 - 09.2023
- Leveraged advanced forecasting modelling techniques to forecast digital conversion rate for all customers, as well as the conversion rate for loyalty customers, in order to monitor the performance of our new loyalty program.
- Developed a calibrated forecasting model which consisted of a Prophet forecasting model, a LGBM forecasting model and a Sarimax model to forecast daily website traffic with relative accuracy to help Merchandize team and Planning team with better planning and budget allocations.
- Applied multiple machine learning algorithms to build a Recommendation Engine for the new Loyalty Program to facilitate customized offers through Email Marketing and new Loyalty Program.
- Leveraged NLP techniques to help Site Operations team design and build an automation system that can extract most relevant keywords from products' descriptions, enable auto-tagging and optimize the filtering function in left navigation bar on the e-commerce website. (Search functionality optimization might be phase 2).
- Guided Rotman students from UoT(2023) with a Product Attribute Extraction project to build a product dictionary using NLP algorithms and techniques.
- Designed the modelling approaches for a hierarchical supply-demand forecasting project, led the project data science team from Marketing and Planning to develop various forecasting models with advanced forecasting techniques for all phases, and built the end-to-end automation pipeline in Dataiku.
- Developed implementable strategies to build an optimization algorithm modeled as a travelling salesman problem(Synthetic Control) after meeting with Marketing leadership, Meta and Google, led Marketing Analytics and CRM teams to code and build the automation to dynamically AB test the ROAS sensitivity for 4 channels with Meta and Google on FSAs.
- Tutored Rotman students from UoT(2022) with a Call Centre Volume Forecasting project, from understanding business requirements, challenges and limitations, getting a good sense of related datasets, to modelling approaches and pipeline creation. And provided technical supports on a weekly basis.
- Mentored data scientists from Planning Analytics team(in India) with their demand forecasting project.
- Provided technical supports to Marketing Analytics team(in India) with their traffic forecasting project.
- Implemented a 2019 vs. 2022 Customer Behavior(pre-Covid vs. post Covid) analysis for Marketing leadership to better understand how our customers' shopping behavior has changed to help them make more targeted strategies towards different cohorts.
- Conducted a Post-Store Closure analysis for stores which were closed during the pandemic and analyzed the impact of the closures on customers, other stores and e-commerce website for the Store team.
- Developed Clustering models in AWS EMR to segment a group of survey customers, and mapped back to the entire customer database with Lookalike models to create valuable and actionable segmentations for Marketing team.
Data Scientist/Machine Learning Engineer/Developer
Pethealth Inc
04.2019 - 06.2021
- Customized the whole machine learning solutions end to end for Full Claims Automation Project, in connection with OCR system and claims portal, without human intervention.
- Conducted time series analysis with sales/conversion forecasting using Statsmodels, Facebook Prophet and RNN to measure the performance of Marketing project, and provide insights for further improvements.
- Built machine learning, deep learning models in Python with various algorithms and advanced techniques, mainly in the area of NLP, customer segmentation and customer retention to help Marketing, Sales and Risk Management teams better understand their leads and customers, and conduct targeted strategies.
- Designed the entire process and work with data engineer to build pipelines in Azure Data Factory and Azure DataBricks to bring customers' data from Google Analytics to Datahub, and develop propensity model with scored segments for Marketing team to do remarketing campaigns.
- Developed machine learning APIs with Django REST framework and Flask, and deployed them to IIS in Azure.
- Integrated machine learning APIs into existing web applications and platforms by changing workflows and added additional logic in the backend with .Net framework.
- Developed affiliated Python applications to help integrated systems working efficiently with machine learning APIs.
- Conducted statistical analysis to help Marketing and Sales team to analyze sales data, discover trends and potential patterns, seek for solutions from different aspects to increase sales performance and conversion rate.
- Created SSRS reports, PowerBI dashboards and automated Excel reports with macro to demonstrate important discoveries and key insights for Marketing and Finance teams.
- Developed new CRM system with latest Microsoft Asp.Net Core 3.1 framework and React front-end.
- Upgraded and rebuilt the current customer web portal from .Net framework 4.5 to Asp.Net Core 3.1.
- Built complicated pipelines in SSIS with heterogeneous data sources , C# programming, stored procedures and VBA macro to help Finance team automate monthly reconciliation process.
- Guided and supported Marketing team and third-party Slalom to design and build the new preference centre in Salesforce Marketing Cloud, and integrated it with all running journeys and automations.
- Projected and developed a whole suit of automation for customer retention and up-sale in Salesforce Marketing Cloud.
- Designed and fixed other automations, such as Gift&Trial, Promo, Newsletter etc in Salesforce Marketing Cloud to help Marketing team implement various email campaigns to the right target groups.
Data Analyst
Riverland Research
03.2018 - 03.2019
- Engaged with stakeholders and clients to elicit business objectives, scope and evaluation criteria (KPI, metrics), remove ambiguous, nail down details, set up assumptions and constraints where apply analytical knowledge.
- Acquired internal and external data from heterogeneous data sources, such as files (txt, csv, excel), FTP, databases, web scraping, and 3rd parties, and migrate data to analytic staging environment.
- Implemented EDA in Python for data investigations so as to discover patterns, spot anomalies, test hypothesis and check assumptions with the help of summary statistics and graphical representations.
- Performed data preprocessing with SSIS or Python, including cleansing, outliner removal, feature scaling, binarization, normalization, encoding according to the nature of different algorithms.
- Conducted data wrangling including data binning, aggregation and other transformations to combine independent variables, calculate ratios, or augment with demo/geo graphic info to boost dataset’s analytic capability.
- Built hybrid models with logistic regression / decision tree for churn analysis, propensity to products, response rate to an email campaign, and predictions on customer shopping behaviors to formulate pre campaign analysis.
- Segmented the existing customers with RFM and clustering model in retailing to analyze customer lifetime value, and provide support for marketing and sales department to customize promotions to target different segments.
- Assessed the performance for each model in term of confusion matrix, error rate, variance deduction, etc, adjust model configuration parameters, and key influencers to improve accuracy and choose the most robust model.
- Analyzed data using advanced SQL queries in SQL server or Tableau, visualized the results with different types of charts to identify patterns, trends, and derived powerful insights from them.
- Summarized the insights with graphs and recommendations in reports, or created dashboards upon requests.
- Deployed and integrated deliverables into the business application for enterprise decision support process.
Business Data Analyst/Commercial Manager
CBMI Construction Co., Ltd
02.2016 - 02.2018
- Used SQL to analyze each sub-contractor’s historical quote data and previous critical contract terms to help account managers make new quotations and prepare deviation lists for new projects.
- Analyzed the changes of cement industry and our competitors to extract insightful information to support strategy making and business operations.
- Prepared interactive monthly progress reports of all projects in construction with Excel VBA.
- Produced MTD reports and quarterly reports of sales performance for marketing department with statistical information and charts generated in tableau.
- Participated in the critical formal business negotiation meeting in the Owner's premise and took charge of business part negotiations with the owner's commercial team.
Commercial Manager
CBMI Construction Co., Ltd
07.2009 - 08.2013
- Identified the expected profit, time schedule, payment terms and liquidated damages with management, scrutinized the commercial tender, sought out the clauses which contain potential risks and ran counter to our expectations.
- Prepared commercial deviation list, itemized the selected clauses with its descriptions in contract and our comments.
- Negotiated with owner on the deviation list, participated in formal business negotiation meeting in Owner’s premise, and took charge of business part negotiations with owner’s commercial team.
- Organized revised commercial documents, meeting minutes and variation orders for handover to relevant departments.
- Assisted and coordinated with contract management department, logistic department, financial department and project department to facilitate the smooth progress of initial stage of the project.
Education
M.D. - Business Administration And Management
Wuhan University Of Technology
Wuhan, Hubei, China06.2016
J.D. - Business Administration And Management
Zhongnan University Of Economics And Law
Wuhan, Hubei, China06.2009
Skills
- Machine Learning and Deep Learning - 5 years
- Large Language Models (LLMs) - 1 year
- LangChain - 1 year
- ChatGPT, Azure OpenAI - 1 year
- Azure App Services, Azure Container Registry, Azure CosmosDB, Docker - 3 years
- FastAPI - 1 year
- Web Scraping (BeautifulSoup, Selenium) - 1 year
- R - 1 year
- Dataiku - 3 years
- Python, Django, Flask - 5 years
- Tensorflow, Keras - 5 years
- PyTorch, PySpark - 3 years
- NLP, Time Series Modelling - 5 years
- Statistical Analysis - 5 years
- Dataiku - 2 years
- AWS SageMaker, S3, EMR, EC2 - 1 year
- Azure Data Factory, Azure ML, Azure DataBricks - 1 year
- Web Application, API Development - 3 years
- C#, ASPNET Core - 1 year
- Snowflake - 2 year
- SQL, SSIS - 7 years
- PowerBI, Tableau, SSRS, Excel VBA - 5 years
Objectives
Senior Machine Learning Engineer / Senior Data Scientist
Accomplishments
At HBC (Hudson's Bay):
- Website Traffic forecasting project which reduced the forecast variance from 40% to 8% compared with the corporate forecast model developed by Finance team, helps Merchandize team and Planning team to make better plannings and budget allocations.
- Supply-Demand forecasting project which significantly improves the accuracy of monthly demand forecasting, and helps the monthly go-to-market process and planning.
- Synthetic Control project (an optimization algorithm) which enables Marketing team to effectively AB test the ROAS sensitivity for 4 channels with Meta and Google on FSA level.
- Nominated by the leadership and represented the Finance sector to participate in the Leadership Program.
- Tutored Rotman students from UoT with a Call Centre Volume Forecasting project from end to end which improves 38% of the accuracy compared with the vendor's forecasting.
At Pethealth Inc:
- Full Claims automation ML project (NLP) which increases approximately 50% of daily efficiency.
- US Trial Portal optimization project which uses time series analysis to forecast the trend and evaluate performance.
- Customer retention and up-sale project with Salesforce Marketing Cloud which encourages new customers to upgrade their insurance products and generates more sales.
- Finance reconciliation automation project which significantly saves the amount of work and time for finance team to do monthly reconciliation.
- New customer portal web development project.
- Awarded twice of company's Recognition Program by North American Leadership.
Timeline
Senior Data Scientist /Machine Learning Engineer
Financial Services Regulatory Authority Of Ontario
09.2023 - Current
Data Scientist /Machine Learning Engineer
HBC ( Hudson's Bay )
06.2021 - 09.2023
Data Scientist/Machine Learning Engineer/Developer
Pethealth Inc
04.2019 - 06.2021
Data Analyst
Riverland Research
03.2018 - 03.2019
Business Data Analyst/Commercial Manager
CBMI Construction Co., Ltd
02.2016 - 02.2018
Commercial Manager
CBMI Construction Co., Ltd
07.2009 - 08.2013
M.D. - Business Administration And Management
Wuhan University Of Technology
J.D. - Business Administration And Management
Zhongnan University Of Economics And Law
Similar Profiles
Rohit SinglaRohit Singla
Cyber Subject Matter Expert, Identity & Privilege at Financial Services Regulatory Authority Of OntarioCyber Subject Matter Expert, Identity & Privilege at Financial Services Regulatory Authority Of Ontario
Warren UrquhartWarren Urquhart
Articling Student at Ontario - MPBSD And MOIArticling Student at Ontario - MPBSD And MOI
Demir B. KiziltanDemir B. Kiziltan
Intern at FSRA (Financial Services Regulatory of Ontario)Intern at FSRA (Financial Services Regulatory of Ontario)