Free datasets

Where to download free datasets — and what to build with them

Every portfolio project starts with real data. These are the sources our mentors use, what each one is good for, and the project we recommend building from it.

16 trusted sources

Dataset download websites

Global repositories plus Gulf open-data portals so your portfolio speaks to local employers.

Kaggle Datasets

Free

Portfolio projects and practice for every skill level

Retail salesChurnHousingHealthcareFinanceImagesText

Mentor tip: Filter by 'Usability > 8' and CSV under 100MB so the data loads in Excel and Power BI without pain.

Open source

Google Dataset Search

Free

Finding a dataset on any specific topic

Everything indexed across the web

Mentor tip: Search the business question, not the file type: 'Kuwait electricity consumption by month'.

Open source

data.gov

Free

Government, economics and public services data

EconomyTransportEnergyEducationClimate

Mentor tip: Great for credible, citable projects — recruiters trust official sources.

Open source

Kuwait Central Statistical Bureau

Free

Kuwait population, labour, trade and price statistics

DemographicsLabour marketCPITrade

Mentor tip: Local data makes your portfolio instantly relevant to Kuwaiti employers.

Open source

Saudi Open Data Portal

Free

Saudi ministries, Vision 2030 indicators and municipal data

HealthEducationTourismEnergyTransport

Mentor tip: Pair a Saudi dataset with a Vision 2030 KPI for a standout dashboard.

Open source

Dubai Pulse / UAE Open Data

Free

UAE smart-city, transport and real estate data

Real estate transactionsTrafficTourismUtilities

Mentor tip: Dubai land transaction data is a superb Power BI project — rich dimensions and real money.

Open source

World Bank Open Data

Free

Country-level macroeconomic time series

GDPPopulationEmploymentTradeDevelopment

Mentor tip: Has a clean API — perfect for practising API ingestion in Python.

Open source

UCI Machine Learning Repository

Free

Classic ML benchmark datasets

ClassificationRegressionClusteringTime series

Mentor tip: Small and clean — ideal for learning algorithms without cleaning overhead.

Open source

Our World in Data

Free

Well-documented global indicators with charts

HealthEnergyClimatePopulationFood

Mentor tip: Study their chart design — it is a masterclass in data storytelling.

Open source

AWS Open Data Registry

Free

Very large datasets for data engineering practice

Satellite imageryGenomicsWeb crawlWeather

Mentor tip: Use these to practise Spark, partitioning and cloud storage costs at real scale.

Open source

BigQuery Public Datasets

Free + Paid

Practising warehouse-scale SQL

GitHub activityStack OverflowNYC taxiWeatherCrypto

Mentor tip: The 1TB monthly free query tier is plenty — always preview with LIMIT and partition filters.

Open source

HuggingFace Datasets

Free

NLP, LLM and multimodal projects

Text corporaQ&ASentimentTranslationAudio

Mentor tip: Load in two lines with the datasets library — great for a generative AI capstone.

Open source

Data.world

Free + Paid

Community datasets with SQL in the browser

BusinessSportsMarketingSocial

Mentor tip: Useful for quick, small business-style datasets when you need a scenario fast.

Open source

FiveThirtyEight

Free

Story-driven datasets behind published articles

PoliticsSportsCulturePolls

Mentor tip: Read their article, then try to reproduce the analysis — excellent interview talking point.

Open source

Maven Analytics Data Playground

Free

Clean, business-shaped datasets built for BI practice

SalesHRMarketingAirlineRestaurant

Mentor tip: Closest to what you will actually meet in a Gulf reporting job.

Open source

Open Data Kuwait / GCC Stat

Free

Cross-GCC comparative statistics

PopulationLabourTradePrices

Mentor tip: Comparing Kuwait, Saudi and UAE on one dashboard shows regional business awareness.

Open source

Hugging Face Datasets

Free

NLP, LLM and computer-vision projects

TextChatImagesAudioTranslation

Mentor tip: Load any dataset in one line with the datasets Python library.

Open source

Maven Analytics Data Playground

Free

Business-style datasets for Power BI and Excel portfolios

RetailHRRestaurantAirlineMarketing

Mentor tip: Each dataset includes a data dictionary — perfect for realistic dashboards.

Open source

Our World in Data

Free

Global development, health and energy analysis

PopulationHealthEnergyEconomyClimate

Mentor tip: Every chart has a downloadable CSV — great for storytelling projects.

Open source

World Bank Open Data

Free

Country-level economic indicators for Pakistan, the Gulf and beyond

GDPInflationEducationEmployment

Mentor tip: Compare Pakistan, Saudi Arabia and UAE side by side for a strong regional project.

Open source

Pakistan Bureau of Statistics

Free

Pakistan census, labour and price data

CensusCPILabour forceTrade

Mentor tip: Use CPI series for an inflation-tracking dashboard relevant to local recruiters.

Open source

data.gov (USA)

Free

US government open data across every sector

TransportHealthEducationCrime

Mentor tip: Great for projects aimed at US and Canada employers.

Open source

Open Government Canada

Free

Canadian federal datasets

ImmigrationEconomyEnvironmentHealth

Mentor tip: Immigration and labour datasets make relevant Canada-focused portfolio work.

Open source

AWS Open Data Registry

Free

Large datasets for data-engineering and Spark practice

SatelliteGenomicsWeatherLogs

Mentor tip: Query directly from S3 with Athena or Spark — no download needed.

Open source

Dataset types

What kind of data can you download — and what to build

Pick the dataset type that matches the job you want, then build the recommended project.

Transactional sales data

Revenue, margin and basket analysis; the backbone of any BI portfolio project.

Build: Star-schema Power BI dashboard with YTD, YoY and store-level drill-through.

Customer / CRM data

Segmentation, RFM scoring, churn prediction and lifetime value.

Build: RFM segmentation in SQL plus a churn model in Python.

Web and app event data

Funnels, retention cohorts, session analysis and A/B tests.

Build: Cohort retention heatmap and a funnel drop-off analysis.

HR / workforce data

Attrition drivers, headcount planning, diversity and salary bands.

Build: HR attrition dashboard with a decision-tree driver analysis.

Financial statements

Variance analysis, budget vs actual, cash-flow trends.

Build: Budget-vs-actual variance report with automated commentary.

Supply chain & inventory

Stock-outs, lead time, demand forecasting and ABC analysis.

Build: Demand forecast with seasonality plus a reorder-point simulator.

Healthcare data

Patient flow, readmission risk, resource utilisation.

Build: Readmission risk model with fairness and privacy notes.

Real estate transactions

Price modelling, area comparisons, yield analysis.

Build: Dubai/Kuwait property price dashboard with a regression model.

Public / open government data

Population, energy, education and transport analysis with credible sources.

Build: GCC comparison dashboard tied to a national strategy KPI.

Text and review data

Sentiment analysis, topic modelling, LLM summarisation.

Build: Customer review sentiment tracker with an LLM insight summary.

Streaming / IoT data

Real-time monitoring, anomaly detection, sensor analytics.

Build: Kafka to warehouse pipeline with a near real-time operations board.

Geospatial data

Catchment analysis, delivery routing, branch network planning.

Build: Map-based branch performance dashboard with drive-time zones.

Before you publish a dataset project

Check the licence before you publish anything — CC0 and CC-BY are safe for portfolios.
Never publish personal data; anonymise names, phone numbers and IDs.
Document the source, download date and row count in your README.
Keep the raw file untouched and do all cleaning in a script or Power Query.
State data-quality issues you found — recruiters value honesty over a perfect story.
Prefer one deep project on a relevant dataset over five shallow ones.

Free learning hub