Business Intelligence Portfolio: Examples, Templates and How to Show a Dashboard
Quick answer: A business intelligence portfolio is two to four dashboards, each presented as a small case study rather than a screenshot. For every dashboard, show who uses it and what they decide with it, the data model behind it (fact and dimension tables, and the grain), two or three of the measures written out in DAX or SQL, the dashboard itself as images plus a link a stranger can open, and the limits of the data. Host Tableau work on Tableau Public; for Power BI, put screenshots and a short video on the page, because the free desktop app runs only on Windows and public web links need an organisation account. If you want analysis projects rather than dashboards, see the data analyst project guide.
Image credit: Mikael Blomkvist / Pexels
How this guide was put together: the facts about Power BI and Tableau Public come from Microsoft Learn and Tableau's help pages, opened on October 21, 2026 and listed at the end. The portfolio statistics come from Seera's benchmark of 56 data portfolios, a data science and analytics sample, not a BI-only one. The three examples are real portfolios built on Seera, read from their public pages on the same day. Seera does not sell Power BI or Tableau and has no relationship with Microsoft or Tableau.
Key Takeaways
A reviewer opening a BI portfolio checks five layers in each dashboard. Most portfolios show only the fourth.
| Layer | What the reviewer is asking | What to put on the page |
|---|---|---|
| 1. User and decision | Who opens this, and what do they do differently? | One line: the role, the moment of use and the decision |
| 2. Data model | Can this person model data, or only drag fields onto a canvas? | A model-view image: fact table, dimensions, grain |
| 3. Measures | Do they know DAX or SQL beyond a sum? | Two or three key measures, written out |
| 4. The dashboard | Does it answer the question at a glance? | Clean screenshots, plus a live link or a short video |
| 5. Limits | Do they know where the numbers stop being true? | Refresh timing, data gaps, what it cannot answer |
Table of Contents
- What is a business intelligence portfolio?
- What should each dashboard in the portfolio show?
- Where should you host Power BI and Tableau work?
- Three real BI portfolios, three ways of showing a dashboard
- Which dashboard projects should you build?
- A case study template and a page template
- Where Seera fits
- Frequently asked questions
- Useful sources and further reading
What Is a Business Intelligence Portfolio?
A business intelligence portfolio is a page that proves you can build reporting other people rely on. That is a different promise from a data analyst's portfolio. An analysis project ends in a finding: "late deliveries lower review scores." A BI project ends in a tool: a dashboard that a sales manager, a dispatcher or a finance team opens every week, trusts, and acts on. The person hiring for a BI analyst, BI developer or reporting role is therefore checking two things the analysis portfolio barely touches: whether your model will hold up when the data grows, and whether a busy non-analyst could use what you built.
That shifts what goes on the page. Fewer projects, each shown deeper. A BI portfolio with three dashboards, each explained down to the table relationships, beats one with ten pretty screenshots. If your work is mostly one-off analysis instead, the data analyst portfolio guide and its annotated example page fit you better.
The gap is visible in Seera's own data. In the benchmark of 56 data portfolios published on Seera, 57% list Power BI and 30% list Tableau, but only 20% list DAX and 21% list Power Query, the two skills that sit underneath a Power BI dashboard. And just one in five link out to a specific project at all. Naming the tool is common. Showing the model and giving the reader something to open is not.
What Should Each Dashboard in a BI Portfolio Show?
Treat every dashboard as a five-part case study. Each part answers one of the reviewer's questions from the table above, and each can be shown in a few lines or one image.
1. The user and the decision
Open with one sentence that names a role, a moment and a decision: "A regional operations lead checks this every Monday to decide which routes need extra buffer time." It tells the reader you designed for a person, which is the core of BI work, and it sets up every chart choice that follows. A dashboard without a named user reads as a tool tutorial.
2. The data model
Show a screenshot of the model view and say in one line what the grain of the fact table is. Microsoft's own modelling guidance for Power BI is built on the star schema, and it puts the idea simply: "Dimension tables enable filtering and grouping" and "Fact tables enable summarization." It also asks that fact tables "always load data at a consistent grain." A reviewer who sees one fact table, four tidy dimensions and a stated grain such as "one row per flight" learns more about you in five seconds than from the whole dashboard. Tableau users can show the same thing with the data source's relationship canvas.
3. The measures
Pick the two or three measures the dashboard depends on and write them out in a code block, with one line on what each one handles. A year-over-year measure that respects the date filter, a ratio that does not average averages, a measure that ignores one slicer on purpose: these are where BI skill actually shows. If you work in SQL rather than DAX, show the view or query that shapes the fact table instead.
4. The dashboard itself
Export clean images of each page: the report canvas only, not the editor around it. Add a live link where you can (see the hosting section below), and a 30 to 60 second screen recording that clicks through the one interaction that matters, such as a drill-through or a filter that changes the story. The recording covers every reader who cannot or will not open the live version.
5. The limits
Close with what the dashboard cannot tell its user. How often the data refreshes, which months are missing, which metric is an approximation. In BI, the person who states the limits is the person who will not ship a misleading number to a director. One short paragraph is enough.
Where Should You Host Power BI and Tableau Work?
This is where BI portfolios most often break. A reviewer clicks a link and either sees nothing, sees a download prompt, or sees a sign-in page. The four options behave very differently:
| Option | What a viewer needs | Catch |
|---|---|---|
| Tableau Public | A browser | Free, but everything is public; workbooks and data can be downloaded unless you switch that off; 15 million rows per workbook and 10 GB of storage |
| Power BI "Publish to web" | A browser | Needs a Power BI sign-in from a work or school account and the setting enabled by that account's admin; anyone can reach the underlying data |
| A .pbix file on GitHub | Power BI Desktop, a Windows app | A reviewer on a Mac or a phone cannot open it |
| Screenshots and a short video | Nothing | Not interactive, so pair it with one of the options above |
Tableau Public
Tableau describes Tableau Public as "a free platform to explore, create, and share data visualizations using publicly available data online," and its help pages are blunt about the trade: "Workbooks and data published to your Tableau Public profile are not private and are freely accessible to anyone." The FAQ adds that you can "toggle off the ability to download the workbook and the underlying data set in the viz settings," that you can hide vizzes on your profile, and that Tableau Public "only supports extracted data sources, and does not support live connections to data." The limits that matter for a portfolio are 15 million rows per workbook and 10 GB of storage. Tableau also states that its free Public Edition of Tableau Desktop "is not for commercial use," which is fine for a portfolio built on public data and a reason to keep client work off it.
Power BI: publish to web
Power BI Desktop is free, but Microsoft lists Windows 10 or later as the minimum requirement, so a .pbix file is not something most recruiters can open. Publishing to the web goes through the Power BI service, and Microsoft's sign-up page states: "To sign up for the Power BI service, you need a work or school email account. You can't use consumer email accounts (for example, Gmail or Hotmail) directly." Its suggested route for people without one is a Microsoft 365 trial that creates a work account, which needs a credit card.
Once you are in, Publish to web also requires that "Your Power BI administrator must enable the Publish to web tenant setting." Microsoft's warning is the reason employers switch it off: "anyone on the Internet can view your published report or visual," and "anyone can access the underlying data in your model even if your report does not display it." Two practical consequences for a portfolio. Never publish a report built on an employer's or client's data this way. And Publish to web does not support every report: Microsoft's list of exclusions includes row-level security, DirectQuery and R or Python visuals.
The setup that works for most people
- Put clean screenshots and a short video on the portfolio page itself, so every reader sees the work.
- Add one live link: Tableau Public for Tableau work, Publish to web for Power BI if you have an account that allows it.
- Put the source file (.pbix or .twbx) and your SQL or Power Query steps on GitHub for the reviewer who wants to inspect the model.
If your strongest dashboards belong to an employer, do not publish them in any of these forms. Describe the scope in words and rebuild the method on public data; the guide to portfolios for confidential work covers how.
Three Real BI Portfolios, Three Ways of Showing a Dashboard
These three portfolios were built on Seera and are live at the time of writing. Each one shows its dashboard work in a different way, and each way has a clear next step. The descriptions come from what each public page shows today.
| Portfolio | Tool | How the dashboard is shown | One next step |
|---|---|---|---|
| Harsha Kumar Pasala | Power BI, Excel, SQL | Each project title links to its source file on GitHub | Add an image of the report to the Power BI card |
| Benedict Chidiebere | Power BI with Python | A screenshot of the report on the project card | Crop the image to the report canvas and add a link |
| P. Tara Bhramaramba | Tableau | A written description of what each dashboard analyses | Publish both to Tableau Public and link them |
The file link: Harsha Kumar Pasala
Harsha Kumar Pasala's portfolio covers Excel, SQL and Power BI with one project apiece, and every project title opens the working file on GitHub. The Power BI project is a mobile sales report delivered as a .pbix file, with a description of the KPIs it tracks. That is the strongest habit in this set: nothing is claimed that cannot be opened. The next step is about who can open it. A .pbix needs Power BI Desktop on Windows, so a reviewer reading on a phone sees only text. An image of the main report page on the card, with the file link kept underneath, puts the work in front of every reader.
The screenshot: Benedict Chidiebere
Benedict Chidiebere's portfolio has a Premier League project that combines machine-learning predictions with Power BI dashboards for investors and business stakeholders. The project card carries a screenshot of the report: transfer spending against commercial revenue by season for three leagues, a ranked table of clubs by commercial revenue, and a chart of each league's share. A reader sees the dashboard without clicking anything, which most portfolios do not manage. The image was captured inside the Power BI editor, so the ribbon and the filter and visualization panes share the frame. Exporting just the report canvas would give the charts the full width, and adding a link or a short recording would let a reviewer try the interactions.
The description: P. Tara Bhramaramba
P. Tara Bhramaramba's portfolio has two Tableau projects. One looks at how house prices move with the features of a property, with renovations and with location. The other is a dashboard concept for rainwater harvesting covering water availability, usage, cost savings and sustainability metrics. Both descriptions say clearly what the dashboard covers, which is the user-and-question layer done in words. Neither has a link or an image yet, and Tableau Public is a free home for exactly this kind of work: publishing both and adding the links would turn two descriptions into two dashboards a reviewer can open.
More data portfolios built on Seera are collected in the data science and analytics showcase.
Which Dashboard Projects Should You Build?
Good BI projects share three traits: a recurring decision someone makes, data that keeps arriving, and enough volume that the model matters. Each project below names the user, the fact table's grain and the measures worth writing out. They use different public sources from the eight analysis projects on this blog, so you can build both kinds without repeating a dataset.
| Project | User and decision | Fact grain | Measures to show |
|---|---|---|---|
| Airline punctuality | An airport operations lead deciding where to add schedule buffer | One row per flight | On-time rate, average arrival delay, cancellation rate |
| Taxi demand by zone | A fleet dispatcher deciding where to position cars each hour | One row per pickup zone per hour (aggregated) | Trips per hour, median fare, share of airport trips |
| Bike-share station balance | A rebalancing team deciding which stations to refill before the morning peak | One row per trip | Net flow per station per hour, share of casual riders |
| Your own job search | You, deciding which application channel deserves your time | One row per application | Response rate by channel, days to first reply, stage conversion |
Airline punctuality (Bureau of Transportation Statistics)
The US Bureau of Transportation Statistics publishes on-time data reported by the 17 US carriers with at least 0.5 percent of domestic scheduled-service passenger revenue, including departure delay, taxi-out time, cancellations and diversions by airport and airline. The modelling point to show: each flight has an origin and a destination airport, so the airport dimension plays two roles. Microsoft's guidance calls this a role-playing dimension and describes two ways to handle it in Power BI. Explaining which one you chose, and why, is a strong paragraph for an interview.
Taxi demand by zone (NYC Taxi and Limousine Commission)
The TLC publishes trip records every month as Parquet files, "typically with a two-month delay," with a taxi zone lookup table alongside. Each monthly file holds millions of trips, so this project lets you show a decision most portfolios skip: aggregating to zone and hour before the data reaches the dashboard. Say on the page what you aggregated and why. It also keeps a Tableau Public version under the 15-million-row limit.
Bike-share station balance (Citi Bike)
Citi Bike's trip histories record start and end times, start and end stations with coordinates, the bike type, and whether the rider is a member or a casual user. Net flow per station by hour is a measure that cannot be built with a single drag-and-drop, which is why it belongs in the measures block. Check the operator's data use policy, which the download page points to, before you publish.
Your own job search
A dashboard of your own applications has two advantages: the data is unambiguously yours, and the decision is real. Log each application with its channel, role, date sent and stage dates, and the measures write themselves. It is also an easy story in an interview, because you were the user.
A note on sample datasets
The datasets bundled with BI tools are good for learning and weak as portfolio centrepieces. Reviewers have seen them many times, and the terms can be narrower than they look: Microsoft's Power BI sample page says "You can copy and use this workbook for your internal reference purposes" and asks that any use carry the data provider's attribution. Practise on the samples; publish on public data you can explain.
A Case Study Template and a Page Template
Copy this block for each dashboard. Every line is one sentence or one image, so a full case study fits on one screen.
Title: [the decision, not the tool], e.g. "Where to add schedule buffer: a punctuality dashboard" Who uses it: [role] checks it [when] to decide [what] Data: [source and date range]; refreshed [how often] Model: [1] fact table at [grain]; dimensions: [list] -> model-view image Key measures: [Measure 1] = [DAX or SQL] // what it handles [Measure 2] = [DAX or SQL] // what it handles Dashboard: [2 to 3 page images]; live link: [Tableau Public or Publish to web]; video: [30 to 60 s] What it shows: [the one finding a user would act on] Limits: [what the data cannot tell you] Files: [GitHub link to .pbix / .twbx and queries]
Filled in for the airline project, the first four lines would read: "Where to add schedule buffer: a punctuality dashboard. An airport operations lead checks it every Monday to decide which routes get extra turnaround time. Data: BTS on-time reporting for the last 24 months, refreshed monthly. Model: one fact table at one row per flight; dimensions for date, carrier, origin airport and destination airport."
And the page around the case studies stays short:
- Headline. The role and the tools in one line: "BI analyst: Power BI, SQL and dimensional modelling."
- Two to four dashboard case studies, each in the template above, strongest first.
- Skills, grouped so the model layer is visible: modelling (star schema, Power Query, DAX), databases (SQL dialects), visual tools.
- Experience, with any work dashboards described in words: the users, the refresh cadence, the decision they supported. No internal screenshots.
- One contact route.
Where Does Seera Fit?
Seera builds an editable portfolio website from the resume you upload. Your skills and experience come across from the resume, and each project card takes a title, a description, the tools used, an image and a link, which is room for a report screenshot plus a live Tableau Public or Power BI link per dashboard. The case study detail from the template above goes into the description. Seera does not host dashboards itself, so the interactive version stays on Tableau Public or the Power BI service. Start from the resume to website tool, and see the data analyst portfolio example for a finished page in this family.
Turn your resume into a portfolio page →
Frequently Asked Questions
What is a business intelligence portfolio?
A business intelligence portfolio is a page of two to four dashboards, each presented as a short case study: who uses it and what they decide with it, the data model behind it, the key measures in DAX or SQL, images of the dashboard plus a link a stranger can open, and the limits of the data. It proves you can build reporting other people rely on, not just make charts.
How many dashboards should a BI portfolio have?
Two to four. Reviewers read a BI portfolio for depth, so three dashboards explained down to the data model beat ten screenshots. Pick projects with different users and different modelling problems, and put the strongest first.
Should a BI portfolio use Power BI or Tableau?
Use the tool named in the job ads you are applying to. In Seera's benchmark of 56 data portfolios, 57% list Power BI and 30% list Tableau. The tools also differ in hosting: Tableau Public lets anyone publish free public dashboards, while a public Power BI link needs a work or school account with Publish to web enabled, so Power BI portfolios lean more on screenshots and video.
How do I share a Power BI dashboard in a portfolio without a work account?
Put clean screenshots of each report page and a 30 to 60 second screen recording on the portfolio page, and link the .pbix file on GitHub for reviewers who use Power BI Desktop. Microsoft requires a work or school email to sign up for the Power BI service, which Publish to web depends on; its suggested route for people without one is a Microsoft 365 trial that creates a work account.
Is it safe to use Power BI Publish to web for portfolio projects?
Only for data you are happy to make fully public. Microsoft warns that anyone on the internet can view a published report without signing in and can access the underlying data in the model, even data the report does not display. Use it for public datasets, and never for an employer's or client's data.
Can I use sample datasets like Superstore or Microsoft's Power BI samples in my BI portfolio?
Use them to practise, not as the centrepiece. Reviewers have seen them many times, and Microsoft's Power BI sample terms allow copying the workbooks for internal reference purposes and ask for the data provider's attribution. Public data with a real recurring decision, such as airline punctuality or bike-share trips, makes a stronger portfolio project.
Related Reading
- 8 Data Analyst Portfolio Projects That Get Interviews
- Data Analyst Portfolio: 3 Project Templates That Get Hired
- Data Analyst Portfolio Example — Annotated Walkthrough
- How to Build a Data Analyst Portfolio With No Experience
- Financial Analyst Portfolio: What to Show When Your Work Is Confidential
- 7 Real Data Science Portfolio Examples Worth Studying
- 10 Best Data Scientist Portfolios — Real Examples
- Data Scientist Skill Benchmarks (2026)
Useful sources and further reading
- Microsoft Learn: Publish to web from Power BI — source of the public-access warning, the tenant-setting requirement and the list of unsupported reports
- Microsoft Learn: Sign up or purchase Power BI as an individual — source of the work or school email requirement and the Microsoft 365 trial route
- Microsoft Learn: Download Power BI Desktop — free download; minimum requirement Windows 10 or later
- Microsoft Learn: Understand star schema and the importance for Power BI — source of the fact and dimension quotes and the role-playing dimension guidance
- Microsoft Learn: What are Power BI samples — source of the sample-use terms quoted above
- Tableau: Tableau Public FAQ — free platform, 10 GB storage, 15 million rows per workbook, download toggle, extracts only
- Tableau: Publish to Tableau Public — source of the statement that published workbooks and data are not private
- Tableau: Desktop edition comparison — Public Edition "is not for commercial use"
- Bureau of Transportation Statistics: Airline On-Time Statistics — data for the airline punctuality project
- NYC Taxi and Limousine Commission: Trip Record Data — monthly Parquet trip files and the taxi zone lookup
- Citi Bike: System Data — trip histories and the link to the data use policy
- Seera: Data Scientist Skill Benchmarks — where the Power BI, Tableau, Power Query, DAX and project-link shares come from; 56 anonymised data portfolios, published under CC BY 4.0
- Seera: Resume to Website — the starting point named in this guide: upload a resume, get a portfolio page you can edit