Urban Eats SQL Analysis
Designed a relational schema and SQL reporting layer for a multi outlet café business.
Read Case StudySydney, NSW 🇦🇺 · Australia-based Data Analyst · Open to remote
Muntasir Md NafisData Analyst Portfolio
Data Analyst · Sydney based · Open to remote
I like the messy part: taking chaotic data and making it undeniable.
Graduate Temporary Visa (subclass 485) · Full unrestricted working rights in Australia
0 Projects0.00M Records AnalysedPower BI · SQL · PythonSydney · Open to Remote
Featured Work
Two projects worth opening first
Hand picked case studies that show how I move from messy data to decisions a business can actually act on.
NEM Carbon Emissions Analytics (2019–2025)
Six years of Australian electricity market emissions mapped across five states to surface clean and dirty windows.
Read Case Study01: About
Bridging code, statistics, and business decisions.
I studied Computer Science, which pulled me into data and a year of machine learning research. The technical side came easily, but I kept hitting the same wall: I could build the model, not always frame the business question behind it. So I pursued a Master of Business Analytics at Macquarie University to close that gap. The projects I've enjoyed most since sit at the intersection of technical rigour and real business stakes: energy market analysis, carbon emissions modelling, retail forecasting. I care less about the tools (Python, SQL, Power BI) than what they let me do, which is turn ambiguous problems into clear recommendations. I'm looking for Data Analyst roles where turning messy data into clear decisions is the job.
Master of Business Analytics
Macquarie University
Predictive Modelling
Churn Prediction · Inventory Risk · Forecasting
Data Visualisation
Power BI · Tableau
SQL & Analytics
MySQL · PostgreSQL · Business Reporting
02: Journey
From Dhaka to Sydney
A decade of learning, research, moving countries, and building toward a career in data.
2018 to 2022
Bachelor of Science in Computer Science and Engineering
North South University, Dhaka, Bangladesh. Built a foundation in algorithms, software engineering, and research methods.
2023
Research Assistant, Machine Learning and Computer Vision
North South University. Prepared datasets, trained models, and evaluated object detection pipelines using YOLOv5 and Detectron2.
2024
Moved to Sydney, started Master of Business Analytics
Macquarie University, North Ryde, Australia. Bridged technical depth with business storytelling, stakeholder communication, and decision science.
2025
Runner up, Business Analytics Capstone Competition
Macquarie University. Recognised for analytical rigour and delivering business impact under tight deadlines.
2026
Launched mmnanalytics.com, Building toward a Data Analyst role in Sydney.
Building a public portfolio of analytics work across energy, retail, and business domains.
2026
Data Analyst Intern, Payreq
Tableau & multi-timezone analytics. Solving data visualization challenges for cross-country reporting; standardizing timestamp formatting and creating timezone-aware KPI dashboards that serve 3 business regions.
03: Skills
Tools & Technologies
Technologies and techniques I use regularly, grouped by domain.
Core
PythonSQLPandasNumPyScikit-learnMatplotlibPower QueryPostgreSQLExcelJupyter
Visualisation & BI
Power BITableauDAXPower QueryReport DesignSeaborn
Analytics
EDAPredictive ModellingFeature EngineeringTime Series ForecastingStatistical AnalysisKPI DevelopmentBusiness Reporting
Database
MySQLPostgreSQLERD DesignRelational Modelling
Business & Communication
Stakeholder ReportingData StorytellingRequirements GatheringAttention to DetailProblem Solving
04: Education
Academic background
Master of Business Analytics
Macquarie University· North Ryde, NSWAchievements: 2nd Place: Business Analytics Capstone Project Competition (2025)
Bachelor of Science in Computer Science and Engineering
North South University· Dhaka, BangladeshAchievements: Cum Laude Distinction
05: Certifications
Licenses & certifications
Understanding Cloud Computing
DataCampIssued Jul 2026
Credential ID: 48,570,070
Cloud deployment models (public, private, hybrid), IaaS/PaaS/SaaS service models, data regulations and PII, and a comparative analysis of AWS, Microsoft Azure, and Google Cloud.
Cloud ComputingMicrosoft AzureAWSGoogle CloudIaaSPaaSSaaS
Show credential Introduction to Power BI
DataCampIssued Jul 2026
Data transformation in Power Query, building interactive reports and dashboards, and applying filters and slicers for business insights.
Power BIDAXPower QueryData ModelingData VisualizationBI
Show credential Joining Data in SQL
DataCampIssued Jul 2026
Credential ID: 354,956
Combining data vertically with UNION to stack rows from multiple tables into a single result set, plus INNER, LEFT, RIGHT, and FULL joins for relational analysis.
SQLINNER JOINOUTER JOINUNIONSet OperatorsSubqueries
Show credential Intermediate SQL
DataCampIssued Jun 2026
Credential ID: 325422
Conditional aggregation, GROUP BY, HAVING, CASE statements, COUNT DISTINCT, and scalar subqueries.
SQLConditional AggregationGROUP BYHAVINGCASESubqueries
Show credential 05: Projects
Analytics, BI & machine learning work
Real world projects across forecasting, predictive analytics, business intelligence dashboards, SQL, and applied machine learning research.
Featured Analytics Projects
Forecasting, predictive analytics, and exploratory business analysis.
0 Years of Data
NEM Carbon Emissions Analytics (2019 to 2025)
PythonPandasTime Series AnalysisData VisualisationMatplotlibCSIRO Emissions API
0+gCO₂/kWh peak detected in VIC evening windows
Analysed six years of Australian National Electricity Market carbon emission intensity data across NSW, VIC, QLD, SA, and TAS to identify high emission peak periods, low emission usage windows, and regional emission drivers.
5 Australian RegionsVIC 31% Emission Share900+ gCO₂/kWh PeakTime Series AnalysisRegional Heatmaps
0 Inventory Records
Dead Stock Prediction Using Machine Learning
PythonPandasNumPyScikit-learnMatplotlibJupyter Notebook
0inventory records modelled for stockout risk
Built an end to end predictive analytics workflow using 3,000 inventory records to identify products at risk of becoming dead stock across stock ageing, monthly demand, inventory turnover, ABC classification, and warehouse movement variables.
Logistic RegressionRandom ForestFeature EngineeringInventory Risk PredictionConfusion Matrix Evaluation
0 Transactions
Dibs Retail Analysis (99K Transactions)
PythonPandasNumPyMatplotlibSeabornEDABusiness Analytics
0retail transactions analysed across 10 variables
Analysed 99,461 retail transactions across customer, product category, payment method, date, quantity, price, and shopping mall variables to identify revenue drivers, customer behaviour patterns, and marketing opportunities.
10 Variables5,024 Price Outliers Treated46.3M TL Highest Age Group Spend49.9% Cash Payment ShareRetail Customer Analytics
Dashboard & Business Intelligence Projects
SQL analytics, data modelling, and interactive dashboards.
0 Years of Data
Victoria Road Crash Analytics Dashboard
TableauPythonPandasGeospatial Analysis
0 yrsof Victorian crash data mapped into one dashboard
Built a Tableau dashboard analysing 5 years of Victorian road crash records (2018 2023), visualising hotspots, severity trends, and risk factors for road safety analysis.
Tableau DashboardGeospatial AnalysisRoad Safety Insights
SQL Analytics
Urban Eats SQL Analysis
SQLSQLiteDatabase DesignERD
Multi tableSQL schema designed for café operations reporting
Designed relational database structures and wrote SQL queries to generate business insights and reporting for retail and café operations.
Database DesignBusiness ReportingData Modelling
Power BI
Australian Population Trends Dashboard (Power BI)
Power BIDAXData ModellingETL
0.00Mpeople mapped across 20 years of ABS census data
Built Power BI dashboards using ABS census data (1996 to 2016) to analyse Australia's population growth, age structure, gender split, and country of birth diversity across all states and territories, supporting planning for infrastructure, healthcare, and education. Found Australia reached 24.18M total population with 8.28% growth, driven primarily by NSW, VIC, and QLD, with 91.11% Australian born and strongest overseas born shares from New Zealand and China.
24.18M PopulationABS Census 1996 2016Demographic Segmentation
Machine Learning Projects
Applied machine learning, computer vision, and deep learning research.
0 Records
Maternal & Neonatal Outcome Prediction
PythonScikit-learnPandasHealthcare Analytics
0birth records modelled across Kenya & Uganda
Predicted adverse maternal and neonatal birth outcomes using machine learning on 61,018 healthcare records from Kenya and Uganda. Compared multiple classification models including Random Forest, Decision Tree, KNN, and Logistic Regression to identify key risk factors and support data driven healthcare decision making.
Random ForestHealthcare AnalyticsPredictive Modelling
0 Records
Mobile Price Predictor
PythonScikit-learnLogistic RegressionKNN
0.00%accuracy predicting phone price tier (LogReg)
Built classification models on 2,000 mobile phone records with 20 hardware features (battery, RAM, camera, resolution, connectivity) to predict price range across 4 tiers: low, medium, high, very high. Compared Logistic Regression and K Nearest Neighbors with grid search hyperparameter tuning, identifying RAM, battery power, and display resolution as the strongest price drivers.
LogReg 97.25%KNN K=11Scikit-learn
Group & Collaborative Projects
Team based business analytics and capstone projects, with my specific contributions highlighted on each card.
Team · Churn Prediction & Customer Analytics Lead
0.00M Records
LuminaTech Lighting: Customer Insights (BUSA8000)
0.00Msales records audited; 40.5% customer churn surfaced
Built churn prediction models on 200,000 resampled records using Logistic Regression, KNN, Naive Bayes, and Random Forest. Random Forest identified recency and order frequency as the strongest churn signals.
EDA & Data CleaningCustomer AnalyticsPython · Pandas
Team · Task 1 Lead: Data Preparation & EDA
0K Listings
Used Car Price Prediction: Kaggle Competition
PythonPandasFeature EngineeringEDA
0K × 38used car listings cleaned across 38 raw features
Team Kaggle competition predicting listed prices for ~12,000 used cars across 38 raw features (engine, drivetrain, dimensions, geography, listing metadata). The team trained and tuned regression models against a held out leaderboard; my workstream owned Task 1: turning the messy raw listings into a clean, model ready dataset and surfacing the price drivers the modelling team built on.
38 FeaturesEDA & Feature EngineeringPython · Pandas
06: Experience
Work & research
Data Analyst Intern (Tableau & Multi-Timezone Analytics)
Payreq· Sydney CBD, Sydney, AustraliaJuly 2026 to Present · Full time
Developing Tableau dashboards for multi-timezone operations. Solving data visualization challenges for cross-country reporting; standardizing timestamp formatting and creating timezone-aware KPI dashboards that serve 3 business regions with conflicting time reference needs.
Research Assistant: Machine Learning & Computer Vision
North South University· Dhaka, BangladeshJan 2023 to Dec 2023 · Full time
Built and evaluated YOLOv5 and Detectron2 object detection models on a 968-image UAV dataset, achieving AP50 of 79.94. Contributed to a peer-reviewed manuscript submitted to Expert Systems with Applications.
07: Contact
Let's build something with data.
If you are a recruiter or hiring manager working on Data Analyst or BI Analyst roles in Sydney, I would welcome a conversation. Reach me at md.nafis08@gmail.com or connect on LinkedIn.