LearnAI home

Showing It ยท Lesson 5

How to Show Your Portfolio to Employers

Where to put it and how to talk about it without overclaiming.

Make it one click away

A portfolio that lives on your laptop is not a portfolio. The work has to be somewhere a stranger can reach in one click, without an account, without a download, and without asking you for anything.

You have a few sensible options and none of them cost money. A shared document with view access for anyone holding the link is the lowest friction and perfectly respectable. A simple one page site works if you already know how to make one. A code hosting site with a clear read me file is right if your project involves code. A PDF attached to an application works too, though it cannot be updated once sent.

Whatever you choose, test it properly: open the link in a private browser window, logged out of everything, on a phone. A link that asks for permission or fails on mobile is worse than no link, because it wastes the reader's goodwill at exactly the wrong moment.

In plain English

Read me:
A plain text file that explains what a project is and how to use it, usually the first thing a reader opens.
Link rot:
When a link that once worked stops working, often because a file moved or sharing settings changed.
Walkthrough:
A short spoken or written tour of your project, from problem to result.
Overclaiming:
Describing your work so it implies more than you actually did.

Check your links every few weeks while you are actively applying. Sharing settings change quietly, and the moment you find out is usually the moment someone tells you they could not open it.

Referencing it on a CV

Do not create a portfolio section that lists five things. One line in the right place is stronger than a wall of links, and readers are scanning.

Put a single link near the top, next to your contact details, and give it a name that says what it is rather than the word portfolio. Then, if the project relates to a role you held, add one line inside that role describing it in plain terms. If it does not relate to any role, a two line projects section near the top is fine.

The line itself should be a fact, not a boast.

โŒ Weak prompt

Prompt

Write a CV line about my AI project that makes it sound impressive.

Output

Spearheaded the development of a cutting-edge AI automation platform delivering transformational efficiency improvements across operational workflows.

Platform is an exaggeration of a workflow. Transformational is unmeasured. A reader who has built anything recognises this immediately, and it makes them doubt the rest of the CV too.

โœ… Good prompt

Prompt

Write one CV line, maximum 25 words, describing this project as a plain fact. State what it does and the measured or estimated result. No adjectives about me, no words like spearheaded or transformational. Here is the project: [description].

Output

Built a prompt and spreadsheet workflow that drafts weekly staff rotas from availability notes, cutting the task from about two hours to twenty minutes.

Short, checkable, and it invites the obvious follow-up question, which is exactly what you want because you can answer it.

Checkpoint

One tested link, one plain factual line, and a walkthrough you can give in under three minutes without notes.

The real test is explaining it

Here is the thing that catches people out. The project is not the evidence. Your ability to talk about the project under questioning is the evidence, and the two come apart the moment you did not really build the thing.

Interviewers will not ask you to recite features. They will ask why. Why that approach and not another. What happened when it went wrong. What you would change. Someone who genuinely made the decisions answers those easily and even enjoys it. Someone who pasted a finished thing into a folder starts hedging within two questions.

So rehearse. Not a script, which sounds rehearsed in the bad way, but the shape of the answers.

Prompt you can copy

Here is a write up of a project I built. Interview me about it as a sceptical hiring manager would.

Ask one question at a time. Start easy, then get harder. Push on:

  • why I chose this approach over an obvious alternative
  • what happened when it produced a wrong result
  • what is still manual and why
  • what I would do differently with more time

After each answer, tell me in one line whether it was specific enough to be convincing, and what was missing.

[paste your write up]

Do that once and you will find the two questions you cannot answer well. Those are the ones you will be asked.

Talking about it without overclaiming

There is a comfortable middle between undercutting yourself and inflating. Aim for it deliberately, because both edges cost you.

Undercutting sounds like it was just a small thing, nothing really. It invites the reader to agree with you. Inflating sounds like I built an AI platform that transformed the process. It invites one specific question that ends the conversation.

The middle sounds like: I had a problem that took two hours a week, I built something that gets it to twenty minutes, one part still needs a human because it made this specific mistake, and here is what I would fix next.

Be careful with the words designed, developed, architected and led when a tool did the heavy lifting. Say what you decided, what you wrote, what you tested and what you rejected. Those verbs are true and they are more interesting anyway.

Where else to put it

Mention the project in your covering message in one sentence with the link, rather than describing the whole thing. If you post about your work publicly, post the write up, not a screenshot with a celebratory caption. And when someone in a conversation asks what you have been doing, having one concrete answer ready is worth more than a long list of courses.

Prompt you can copy

Turn this project write up into three versions, all plain and factual:

  1. One sentence for a covering message, including where the link would go
  2. Twenty five words for a CV line
  3. A spoken walkthrough of about ninety seconds, structured as problem, what I built, what went wrong, what I would change

Use only facts from the write up. No adjectives about me. Do not claim any result the write up does not state.

[paste your write up]

Then say the ninety second version out loud, once, to an actual person or an empty room. Reading it silently does not count, and you will hear the sentence you cannot defend the moment you say it.

Prompt you can copy

Here is how I plan to describe my project out loud. Flag every phrase that claims more than the facts support, and every phrase that undersells what I actually did. Give me a plainer replacement for each.

Then tell me the single question an interviewer is most likely to ask after hearing this.

[paste your walkthrough]

Nothing here guarantees an interview. What it does is make sure that when you get one, the strongest thing about you is in the room with you.

๐Ÿ“ Quiz

Question 1 of 4

You have hosted your project write up online. What should you do before putting the link on an application?

Found this useful? Pass it on.