How to Build a Data Portfolio Website

A data portfolio website is not essential. A strong resume and a tidy GitHub profile already get many people hired. A site earns its place when it does something GitHub cannot: present your projects as short case studies that a hiring manager, including a non-technical one, can read in a few minutes. If you build one, build it for that reader.

When a website adds value

• You are changing careers and need to tell a coherent story about why your projects fit a new field.
• Your best work is visual, such as dashboards or analyses with charts, which read better on a page than in a repository.
• Your audience includes non-technical people, such as hiring managers in business teams, who will not open code.
• You write. A site gives your write-ups a permanent home.

If none of these apply, spend the time on another project instead; a better project improves your resume, GitHub and interviews at once.

The pages you need

Keep it small. Three or four pages are enough:

1. Home: one sentence about who you are and what you do, and links to your best two or three projects.
2. Projects: one case study per project.
3. About: a short professional background, the roles you are looking for, and a photo if you like.
4. Contact: email, LinkedIn and GitHub.

A single well-structured page that contains all four sections works too.

A case-study format for each project

Write each project for someone who will not run your code:

• Headline with the result: "Forecasting weekly demand for 50 products with 12% lower error than the existing method."
• The problem: who needed what, and why it mattered.
• The data: where it came from, its size and its main problems.
• What you did: the approach in plain language, including one decision you are proud of and one that did not work.
• The result: one clear chart and the numbers that matter.
• Limits and next steps.
• Links: to the code on GitHub and, if relevant, a live dashboard.

Aim for 300 to 500 words per case study. Reviewers skim, so put the result first and use headings.

Hosting it for free

For a simple portfolio, static hosting is enough: plain HTML and CSS, or a static site generator, with no server or database to maintain. GitHub Pages is a common choice for this. It is a static site hosting service that publishes HTML, CSS and JavaScript from a repository, and it is available for public repositories on GitHub's free plan (GitHub Docs: GitHub Pages).

Its documented limits are generous for a portfolio: a published site may be up to 1 GB, with a soft limit of 100 GB of bandwidth a month and 10 builds an hour. GitHub also states that Pages is not intended for commercial sites or sensitive transactions (GitHub Docs: Pages limits), which a personal portfolio is not. Keep large datasets out of the site; link to their source instead.

A custom domain with your name is a small cost and looks more professional than a default address, but it is optional.

Embedding dashboards and interactive work

If your best project is an interactive dashboard, show it, but carefully. A screenshot with a short explanation of what the viewer should notice loads instantly and always works; a link to the live version lets interested reviewers explore. Embedding a full interactive dashboard directly in the page can be slow and sometimes requires the viewer to sign in to a vendor's service, which many will not do. For notebooks, publish a rendered version rather than asking reviewers to install anything. Whatever you embed, test it from a private browser window on someone else's device before you share the link, because what works when you are signed in may fail for everyone else.

Design: simpler than you think

Hiring managers judge content, not animations. Use a clean template, one readable font, plenty of white space and charts large enough to read. Make sure the site works on a phone, since many people will open your link from one. Avoid long loading screens, background videos and anything that hides your projects behind clicks.

A pre-launch checklist

• Every link works, including GitHub and live dashboards.
• Each case study starts with the result.
• Charts have titles, labelled axes and sources.
• No confidential employer data or client names appear anywhere.
• The site reads well on a phone.
• Your contact details are correct.
• A friend outside data can explain one of your projects back to you after reading it.

Keep it alive

Update the site when you finish a project, and remove weaker work as better work arrives. A portfolio last updated three years ago suggests you stopped learning. Link the site from your resume and LinkedIn, and keep the same projects and wording consistent across all three.

For choosing what to showcase, see portfolio projects that impress recruiters. For the repository side of each project, see GitHub profile tips.

Case-study-worthy projects are built step by step in the Data Analytics course.

Sources

• What is GitHub Pages? (GitHub Docs) — GitHub
• GitHub Pages limits (GitHub Docs) — GitHub