Skip to content
All articles

How to create a portfolio with AI without making things up

Create a portfolio with AI by choosing real work, drafting clear case studies and checking every claim before you publish or share it.

DIT Team 6 min read

A person arranging blank project cards beside a laptop, illustrating an accurate AI-assisted portfolio

To create a portfolio with AI, start with real projects, choose the work that best shows your skills and ask the assistant to help explain your process clearly. Check every detail against your own records, get permission to share other people's work, and never let AI invent clients, results or responsibilities.

If you are new to AI, join the free AI 101 class. It covers ChatGPT, Claude and Gemini for real tasks, teaches a reusable prompt framework and shows where AI answers need checking, with no coding needed.

Decide what your portfolio should show

A portfolio is a selected set of work that helps someone understand what you can do. It might be a simple document, a collection of project pages or a website. You do not need to include everything you have done. Choose examples that are relevant to the opportunity, course or audience you have in mind.

For each project, collect the original materials you can safely use: your notes, drafts, screenshots, sketches, code or finished work. Write down what you personally did, what the brief asked for and what changed during the process. If a project was done with a team, make your contribution clear instead of presenting everyone's work as your own.

Choose examples you can explain

Pick a few pieces that show different skills or stages of your work. A small number of well-explained projects can tell a clearer story than a long gallery with no context. Include practice projects, coursework or personal experiments when they represent your abilities honestly.

Before using a project, check whether it contains private information, client material, unpublished work or another person's contribution. Get permission where needed, remove details that should not be public, or describe the project at a high level without exposing the underlying material.

How to create a portfolio with AI step by step

1. Make a project inventory

List each project, your role, the skills it demonstrates and whether you have permission to show it. Mark details you need to verify. This helps you choose examples based on evidence instead of asking AI to decide what looks impressive.

2. Build a case study from real notes

For each selected project, collect a short brief, the steps you took, the decisions you made and the finished result. Ask AI to organise those notes into sections such as context, approach, contribution and outcome. If you do not have a verified outcome or metric, leave it out rather than making one up.

3. Improve clarity without changing the facts

Ask the assistant to make your writing easier to follow and to point out where a reader might need more context. Compare the revised version with your notes. Restore any important detail that got removed and correct any sentence that overstates your role or claims a result you cannot support.

4. Choose a format that you can maintain

A portfolio can be a document or a website. Pick the format that fits your purpose and the time you have to update it. If you choose a site, start with a small structure: a short introduction, a few project pages and a way for the intended audience to contact you if appropriate.

5. Review the full experience

Check spelling, links, image permissions, mobile readability and any personal details. Ask someone who does not know the project to read one case study and tell you what they understand about your role. Fix unclear sections before you share the portfolio.

A portfolio review conversation between a young professional and a mentor

Try this now: draft one honest case study

Use this prompt with notes about your own work. Remove private information and do not include material you are not allowed to share:

Help me organise a portfolio case study from the notes below. Use only the facts provided. Create sections for the project context, my specific contribution, the process, decisions and verified outcome. Do not invent clients, team size, tools, dates, metrics or results. If information is missing, ask questions or mark it as unknown. Keep the wording clear and concise, and flag any sentence that may overstate my role. Notes: [paste your own approved project notes].

Read the draft beside your original notes. Correct any invented detail, confirm that the work can be shown and make sure the description distinguishes your contribution from the team's. If there is no measured result, explain what you made or learned without pretending there was one.

Make your portfolio easy to review

Use consistent headings so a reader can scan each project. Explain the problem or brief in a sentence, then describe your approach and your role. Include only images or files that help a reader understand the work. Add context to a visual where the image alone might be unclear.

Avoid design choices that make the portfolio harder to use. Keep text readable on a phone, use descriptive link labels and check that important pages load correctly. If the portfolio includes code or a prototype, say what is complete and what remains experimental. Do not imply that an unfinished demo is a production-ready product.

If you want to make a simple portfolio website, DIT's Vibe Coding with AI workshop costs ₹99 + GST and runs live on Zoom on Tuesdays from 7 to 10 pm IST. For a practical introduction before choosing a tool, read what vibe coding means.

Common portfolio mistakes to avoid

  • Claiming work you did not do: state your role and collaborators clearly.
  • Inventing impact: use only outcomes you can verify; leave unsupported numbers out.
  • Publishing private material: get permission or remove identifying details before sharing.
  • Showing too much: choose relevant examples and explain them well.
  • Letting AI erase your voice: edit the draft so it still sounds like you and reflects your actual decisions.
  • Sharing an unfinished demo as a finished product: label experiments and limitations honestly.

Frequently asked questions

Can AI write my portfolio for me?

AI can help organise your notes, improve clarity and suggest questions a reader may have. You remain responsible for the accuracy, permissions and final wording.

What if I have no professional projects yet?

You can include coursework, personal experiments or practice projects if you label them accurately. Explain the brief, your process and what you learned without presenting practice work as client work.

Should I include results or numbers in every case study?

Only include results you can verify and explain. If you did not measure an outcome, describe the work you completed or the decision you made instead of inventing a metric.

Keep learning with a free class

Join the free AI 101 class to practise clear prompts with ChatGPT, Claude and Gemini, with no coding needed. Bring non-sensitive notes from a project and learn how to review an AI-assisted draft before sharing it.

DIT team

Questions about a class?

Dates, payments, Zoom links, which workshop fits you. The team answers on WhatsApp through the day.

+91 92216 36113

The team can help with

  • Choosing a workshop
  • Dates & timings
  • Payments & invoices
  • Zoom & WhatsApp links
  • Syllabus PDFs
  • Corporate training
Ali Mehdi Hemani, Founder & lead trainer at DIT

Nobody else does this

Talk to the founder. Directly.

Most institutes keep their founder behind a sales team. We are sure enough of what we teach to give you Ali's own number. Big question, serious concern, or you just want to know if this is right for you: message him.

+91 80808 77860

Ali Mehdi Hemani · Founder & lead trainer. Please WhatsApp first to check he's free, then call.

Experience centres

Two centres in Mumbai. Classes live on Zoom.

Most students learn online, from across India and from Dubai, Toronto, Australia and many more countries. Drop in to a centre for hybrid sessions and the occasional offline workshop.