GWC Workshop: AI as Your Portfolio Partner!

I’m a data science major, a freshman in college, and a member of Girls Who Code. As a part of their Upskill series, they host workshops for Girls Who Code members to gain the technical skills necessary to reach their goals in STEM. On March 3, 2026, I attended their “AI as Your Portfolio Partner” Worrkshop!

Notes

From Chatbot to Builder: How to “Vibe Code” Your Own AI Career Decision Engine

“If you’ve been using AI primarily as a high-powered search engine or a document summarizer, you’re only scratching the surface of what’s possible… Currently, about 85% of people use AI strictly as a conversational chatbot…]But the real professional advantage lies in moving up the pyramid—from level one (chatting) to level three: building bespoke, small-batch software designed just for you.”

In a recent workshop, we moved past the chat window and into “vibe coding”. This is the process of building functional software by focusing on the “vibe” and the logic of the problem rather than manually writing the source code. Our goal was to create a Career Decision Engine: a custom, locally-hosted web app that evaluates any job opportunity against your unique professional DNA.

“Vibe coding is a way of building software without actually writing software code… I still do not know how to code. But I have developed a framework of how I, as a non-engineer, can effectively build things” (Bethany Crystal).

The “Popit” Framework: Identifying AI-Shaped Problems

Before you open a code editor, you need a strategy. The Popit framework helps you move from a tool-first mindset to a problem-first mindset.

  • P – Problem: Define the hurdle. Instead of asking “What tool should I use?”, ask “What is the career choice I’m trying to clarify?”
  • O – Output: Define exactly what the AI should give you. Do you want a 0–100 fit score, a rubric of non-negotiables, or a list of interview questions?.
  • P – Prompt: These are the specific instructions that drive the AI. Interestingly, once you define the problem and output, the AI can often write a better prompt for itself than you can.
  • I – Input: This is your “Context Pack”. Every piece of digitized info you have—resumes, essays, or even a 20-year plan—is data the AI can use to understand you.
  • T – Test: AI rarely gets it perfect on the first try. You must test the output against a real scenario to see if the logic holds up.

Sprint 1: Building Your Context Pack

The AI is only as insightful as the data you provide. To start, create a folder on your computer and gather 3–5 files that define your professional identity:

  • Your Resume/CV: To establish your hard skills and history.
  • A Personal Bio or LinkedIn PDF: To give the AI a sense of your “voice” and narrative.
  • A Career Goals Document: This is crucial. Define your “North Star,” your top 3 skills to build, and your “non-negotiables”.

Sprint 2: Engineering the Logic

Using a tool like Claude Code and/or a text editor like Cursor, you can ask the AI to synthesize these files into a structured “Professional Profile”. Once the AI understands your background, you put it into Planning Mode.

In this phase, we instructed the AI to build a logic engine that compares a job description to our goals and outputs:

  1. A Weighted Fit Score: Categories are weighted based on what matters most to you.
  2. Trade-off Analysis: What are you giving up if you take this role?
  3. The “Investigate” Flag: If the job description is missing key info, the AI flags it for you to ask about.

For example, we tested an external affairs role against the profile of Nancy Drew. The engine correctly flagged it as a poor fit (65/100) because her investigative skills didn’t align with government affairs.

Sprint 3: The Front-End “Small Batch” App

The final evolution is moving out of the command line and into a real interface. We asked the AI to build a local web app using Node.js and Express. By connecting an API key—stored safely in a .env file to protect your security—we turned a chat thread into a functional tool.

The result? A visual dashboard where you can paste a job link and get a visual analysis of how that role fits into your long-term career path.

Points of Inspiration: The Maker Mindset

Moving from a consumer to a maker requires a psychological shift. Here are the key takeaways for anyone starting their AI journey:

  • You Don’t Need Permission: “I grew up in an era where software was sort of handed to us… The last 18 months, I’ve been rewiring my brain to think… I can design my own functioning app.”
  • The Power of “Small Batch”: You don’t need a million users to make something meaningful. “Small batch software is one of the biggest wins of AI. We no longer need to wait for a SaaS tool to create an app for us.”
  • The Reflexive Habit: “Getting into the reflexive habit of working with AI to help you make decisions is probably the most important thing you could do for your career right now.”
  • Stay Curious: “Every time I build an app, it teaches me about how to build an app better.”

Final Thoughts

Thank you to the Girls Who Code Team for this wonderful learning experience via their webinar series, and Bethany Crystal for providing us with her knowledge.