Curriculum Vitae

Hi, I'm Nijiati.

Data Scientist Biochemist Indie Builder

Biochemist turned data scientist — I build and ship ML systems for healthcare and education, from brain-tumor methylation pipelines to indie ed-tech in daily use.

Skills

Python R SQL Java HTML
scikit-learn pandas NumPy PyTorch Keras matplotlib
Bioinformatics Biostatistics Predictive modeling ETL Pipelines
Jupyter MS SQL Server VS Code Django
English French Chinese Uyghur

Master of Data Science

University of British Columbia Okanagan

GPA 95.3  ·  Scholarship 5000

Focus Python/R , Machine Learning, Deep Learning, Statistical Analysis, Bayesian Statistics

  • Lung Disease Classifier (Keras, deep learning)
  • Diabetes Prediction Model(scikit-learn, healthcare analytics)
  • Milk Quality Prediction (scikit-learn, predictive modeling)
OutcomeCapstone shipped a deployable PHQ-9 prediction model for GreenspaceHealth from 140K+ EHR records; course projects became standalone case studies.

Master of Science in Biochemistry

University of British Columbia Okanagan

GPA 92  ·  Scholarship 24000

Focus Bioinformatics, Experimental Design, Scientific Writing

  • Thesis: The Effects of Lipid Diets on Gut Microbiota
OutcomeThesis on lipid diets and gut microbiota led to a first-author Nutrients (2019) paper and seeded six peer-reviewed publications.

Bachelor of Science

University of British Columbia Okanagan

GPA 84.4  ·  Scholarship 17000

Focus Medical Concentration

OutcomeMedical-concentration foundation that supported the move into bioinformatics and health ML.

Data Scientist

AGEL Assurance (Montreal)

WhatInsurance brokerages spread data across carrier portals and Applied Epic, with no unified view of customers, policies, or claims.
HowLead the integration syncing carrier and Applied Epic data into one source of truth, and built/deployed the corporate website (Django + PostgreSQL on Render). Work primarily in French with the Montreal team.
OutcomeReplaced manual reconciliation with a unified data layer and shipped a website on an extensible Django stack.

Contract Research Associate — Brain Tumor Methylation

Mamatjan Lab — Thompson Rivers University

WhatGlioma subtypes have very different survival, but standard methylation testing is slow and costly.
HowBuilt a leakage-controlled ML pipeline: an XGBoost classifier compressing 404K features to 70 (97–99% accuracy) and a Cox survival model (C-index 0.796) that found novel prognostic loci.
OutcomeA compact panel that could make subtyping faster and cheaper; manuscript in progress with the Mamatjan Lab.

Health Research Data Analyst

University of Calgary

WhatA multi-PI microbiome project needed reproducible analysis an MD, PhD, and dietitian could all use.
HowBuilt shared, automated R/Python workflows and ran db-RDA analysis linking microbiome composition to demographic variables.
OutcomeReplaced ad-hoc scripts with reusable pipelines and turned the stats into actionable insights for the clinical team.

Machine Learning Data Scientist (Capstone)

GreenspaceHealth

WhatDepression severity is buried across EHR data and only scored at a separate PHQ-9 visit, so patients are flagged late.
HowLed the MDS capstone: processed 140K records across 14 SQL tables and trained Random Forest, neural, and XGBoost models, capturing 78% of PHQ-9 variance.
OutcomeShowed PHQ-9 severity can be predicted from routine EHR data; delivered a deployable model and reports to GreenspaceHealth.

Research Data Consultant

University of British Columbia Okanagan

WhatGraduate researchers needed Python and R help but had no central resource.
HowRan Python/R workshops for 30+ researchers plus one-on-one help on modeling, wrangling, and visualization.
OutcomeCut analysis time 40%+ and improved publication quality, with a reusable training curriculum.

Biochemistry Laboratory Technician

University of British Columbia Okanagan

WhatA teaching lab needs working equipment, stocked reagents, and trained TAs, or experiments fail.
HowRan day-to-day lab operations for five years — animal care, inventory, procurement, TA training, and protocol testing.
OutcomeKept teaching labs running without disruption; the inventory and procurement systems outlasted my tenure.
2025

Multimodal Gene Expression and Methylation Profiling Reveals Misclassified Tumors Beyond Histological Diagnosis

Electronics · Co-author · ML pipeline · DOI ↗

Mamatjan Y, Abulizi N

Multi-modal ML pipeline integrating gene expression and DNA methylation that flags tumors misclassified by traditional histology.

2025

The adoption of a westernized gut microbiome in Indian Immigrants and Indo-Canadians is associated with dietary acculturation

npj Biofilms and Microbiomes · Co-led microbiome analysis and visualization pipeline (Python / R) · DOI ↗

D'Aloisio LD, Ballal M, Ghosh S, Haskey N, Abulizi N, et al.

Co-led the gut microbiome analysis and visualization pipeline that connected dietary acculturation patterns to compositional shifts in the gut microbiome of Indian immigrants.

2024

A Gut Microbiome Transition: The Journey from India to Canada

Journal of the Canadian Association of Gastroenterology · Statistical comparison lead · DOI ↗

D'Aloisio LD, Haskey N, Abulizi N, et al.

Designed and executed the statistical comparison of pre- and post-migration microbial communities.

2019

Gut Mucosal Proteins and Bacteriome Are Shaped by the Saturation Index of Dietary Lipids

Nutrients · First author · DOI ↗

Abulizi N, Quin C, Brown K, Chan YK, Gill SK, Gibson DL

Ran the bioinformatics workflow on host gut mucosal proteomic and microbiome data.

2017

Linking the Gut Microbial Ecosystem with the Environment: Does Gut Health Depend on Where We Live?

Frontiers in Microbiology · Statistical analysis (R) · DOI ↗

Tasnim N, Abulizi N, Pither J, Hart MM, Gibson DL

Contributed environmental-driver statistical analysis in R.

2016

Dietary Lipid Type, Rather Than Total Number of Calories, Alters Outcomes of Enteric Infection in Mice

The Journal of Infectious Diseases · Proteomic experiments + analysis · DOI ↗

DeCoffe D, Quin C, Gill SK, Tasnim N, Brown K, Godovannyi A, Dai C, Abulizi N, Chan YK, Ghosh S, Gibson DL

Ran proteomic experiments and contributed to downstream data analysis.

Excel Workshop Instructor

University of British Columbia Okanagan

WhatResearch staff needed targeted Excel and Power Query skills without a full course.
HowRan tailored, project-specific sessions plus self-paced videos using real data.
OutcomeUnblocked researchers in hours; videos extended reach beyond the live sessions.

Courses

  • 1. Excel 2. Power Query

Undergraduate Teaching Assistant

University of British Columbia Okanagan

WhatFirst- and second-year STEM courses are gatekeepers, but instructors can't give every student lab help.
HowTA'd seven courses across biology, chemistry, physics, and biochemistry — labs, tutorials, and exam reviews.
OutcomeSupported the full first/second-year STEM core; reappointed across departments.

Courses

  • 1. BIOL 116, Biology for Science Majors I (2014 Summer Term 2)
  • 2. PHYS 122, Introductory Physics II (2013/2014 Winter Term 2)
  • 3. BIOC 393, Biochemistry Laboratory (2013/2014 Winter Term 2)
  • 4. CHEM 111, Principles of Chemistry I (2013/2014 Winter Term 1)
  • 5. PHYS 111, Calculus-Based Physics I (2013/2014 Winter Term 1)
  • 6. PHYS 102, Electricity, Light, and Radiation (2012/2013 Winter Term 2)
  • 7. PHYS 112, Introductory Physics I (2012/2013 Winter Term 1)

Sessional Lecturer

University of British Columbia Okanagan

WhatMDS students arrive from many backgrounds and must write production Python and R within weeks.
HowTaught Programming in Python and R to 37 MDS students, emphasizing best practices over rote syntax.
OutcomeStudents reached confident pandas/dplyr use within the term; materials reused in later cohorts.

Courses

  • Programming in Python and R for Data Science

Graduate Teaching Assistant

University of British Columbia Okanagan

WhatLarge biology courses leave students without one-on-one help, which is where understanding happens.
HowLed labs, tutorials, and discussions across four biology courses and graded with substantive feedback.
OutcomeSupported hundreds of majors through foundational courses; repeatedly assigned high-enrolment sections.

Courses

  • 1. BIOL 125, Biology for Science Majors II (2015/2016 Winter Term 2)
  • 2. BIOL 116, Biology for Science Majors I (2015/2016 Winter Term 1)
  • 3. BIOL 133, Anatomy and Physiology (2014/2015 Winter Term 2)
  • 4. BIOL 131, Anatomy and Physiology (2014/2015 Winter Term 1)

The Effects of Lipid Diets on the Gut Microbiome II

UBC Okanagan Biology Graduate Symposium

  • Won 2nd Place Poster

Omics Methods to Study the Effects of High-Fat Diets on Gut Bacteria

Bioinformatics Guest Lecturer

  • Guest Lecturer

The Effects of Lipid Diets on the Gut Microbiome I

UBC Okanagan Biology Graduate Symposium

  • Won 1st Place Poster

Techniques in Immunology

Immunology Guest Lecturer

  • Guest Lecturer
2023 Master of Data Science Graduate Program Entrance Scholarship CAD $5000
2016 Graduate Dean’s Thesis Fellowship CAD $6000
2015 UBC Okanagan Graduate Fellowship CAD $6000
2015 BC Proteomic Network Training Fund CAD $4000
2015 Best Poster Award (Health Science MSc) CAD $150
2015 The Golden Key Graduate Scholar Award USD $10000
2013 International Student Faculty Award ( CAD $5000
2014 UBC Okanagan Graduate Fellowship CAD $12000
2012 UBC Okanagan International Community Achievement Award CAD $3000
2012 International Student Faculty Award CAD $5000
2012 International Student Faculty Award CAD $5000
2012 Deputy Vice-Chancellor Scholarship for International Students CAD $500
2011 UBC Okanagan International Community Achievement Award CAD $3000
2011 Deputy Vice-Chancellor Scholarship for International Students ( CAD $1000
2011 Dean’s List

Data Mentorship

University of British Columbia Okanagan

Hours 200

WhatGrad students often have flawed stats or code that won't produce publication-quality figures, and supervisors can't help each one.
HowMentored Master's and PhD students in R/Python and statistical interpretation for journal submissions (2015–2024).
OutcomeMultiple mentees published with these analyses; several now lead their labs' stats workflows.

YouTube Content Creator

BitSize Learning Channel

Hours 40

WhatMost coding and language tutorials are hour-long lectures, with little that's short and finishable.
HowCreated short-form videos teaching one concept in 60–120 seconds for the BitSize Learning channel.
OutcomeBuilt the content library behind the BitSize brand and its daily-lesson product.

Education Leader

Crohn’s and Colitis Kelowna Chapter

Hours 180

WhatPatients in smaller cities lack specialist communities; local awareness has to be built by volunteers.
HowLed education talks, distributed materials, and organized the chapter's annual Gutsy-Walk fundraiser.
OutcomeRaised local visibility and ran a successful fundraiser supporting chapter programs.

Literacy Volunteer

Kelowna Project Literacy

Hours 260

WhatNewcomers and seniors often have no one to teach English or digital skills at their own pace.
HowTaught English one-on-one to ESL learners and ran digital-literacy workshops for seniors over six years.
OutcomeHelped dozens with practical English for jobs and daily life, and got seniors online.

Outreach Volunteer

University of British Columbia Okanagan

Hours 450

WhatThe public, especially school kids, rarely sees university science up close.
HowTrained volunteers in animal handling and taught biology to visiting school groups at UBCO outreach.
OutcomeSustained the program for six years and brought hands-on biology to students who'd never enter a lab.

Volunteer Instructor

Parkinson Recreation Center

Hours 150

WhatMost fitness and cooking classes assume able bodies, leaving adults with disabilities few options.
HowCo-led adapted dance and cooking classes, adjusting activities weekly to each participant's ability.
OutcomeParticipants built skill and confidence; the adapted formats were reused by later instructors.