A portfolio lets an interviewer see your work before they meet you. For freshers with no job experience, it is one of the strongest signals you can send. GitHub is the most common place to host it.
What to include
Aim for three to five projects that together show:
- SQL analysis on multiple related tables.
- Python data cleaning and exploration with pandas and charts.
- A dashboard in Power BI or Tableau, with screenshots.
- Optionally, a machine learning model if you are targeting data science roles.
Variety in industries also helps: one banking, one retail, one healthcare project shows you can learn a new domain.
Structure of a good project repository
project-name/
โโโ README.md
โโโ data/ (or a link, if the data is large)
โโโ sql/ queries, numbered in order
โโโ notebooks/ analysis notebooks
โโโ dashboard/ .pbix file and screenshots
โโโ report.pdf optional short write-upUse clear file names like 01_clean_orders.sql instead of query_final_v2.sql.
The README is the most important file
Most visitors only read the README. Structure it like this:
- Title and one-line summary.
- Business problem: who is asking and what decision they need to make.
- Data: tables, rows, time period, and any issues found.
- Approach: the steps you took, briefly.
- Key findings: three to five bullet points, with numbers from your analysis.
- Recommendations.
- Tools used.
- Screenshots of the most important charts or dashboard page.
Keep it short enough to read in two minutes.
Common mistakes
- Only tutorial datasets. Titanic and Iris show you followed a course, not that you can handle business problems.
- No README, or a one-line README.
- Notebooks with errors or no outputs. Run everything top to bottom before you push.
- Dozens of tiny, unfinished repositories. A few complete projects are better.
- Copied work. Interviewers ask detailed questions; you must be able to explain every step.
Pin your best work
GitHub lets you pin repositories to your profile. Pin your best three or four, and add a profile README with a short introduction and links to each.
Where to find portfolio-worthy problems
You need business context and messy data. Our live projects provide both, with a fictional company and a full requirements document, and you can export your work straight to GitHub. Then add the projects to your resume.
Practise what you just read
Reading helps; writing the code yourself is what sticks. Start here: