Sources

    Is This Degree Worth It?

    AI PRODUCT · CASE STORY

    Is this degree worth it?

    Why an education app should answer the six-figure question before it tracks a single deadline.

    4 min

    Imagine buying a house from the photos alone.

    You'd want the inspection, the neighborhood, and what it's worth in ten years.

    Most people choose a degree from the photos: the name, the brochure, a gut feeling.

    Most education tools help you apply.

    Very few help you decide whether to.

    Meet Nadia. She's ten years into her career and thinking about a part-time master's, or maybe an AI certificate.

    She has forty tabs open. Every program promises "career outcomes."

    Nadia

    Which of these would actually move my career?

    Rule 1

    Help her decide before you help her track.

    A TRACKER SAYS

    36 programs you could apply to

    Ranked by your odds of getting in.

    A DECISION ENGINE SAYS

    8 that advance your goal

    Ranked by what they'd do for you.

    Tracking deadlines is the last step, not the first.

    Rule 2

    Match on what she wants, not what she qualifies for.

    Early version

    #1 match

    First-year undergraduate admission

    She meets every requirement, because it's built for 18-year-olds.

    Qualifying isn't the same as wanting. The app first works out what she's after (a part-time master's, a certificate), then ranks only those.

    Rule 3

    Match on meaning, not keywords.

    UX

    =

    Human-computer interaction

    =

    Interaction design

    Different words, one idea. A keyword search sees three things. A person sees one.

    Rule 4

    The AI proposes. Code confirms.

    AI reads the page

    "GRE 320 required"

    →

    Code checks the page

    Not there

    →

    Result

    Thrown out

    A fact only gets shown if it's really on the source page, and then it's shown with that source.

    PRINCIPLE

    Don't let the AI be the last word

    Language models can state made-up facts confidently. Put a check between what the AI says and what the person sees.1

    Rule 5

    Confidence is earned, not typed in.

    Tuition

    source: program website

    High · official site, checked recently, word for word

    Low · not found on its cited page

    How sure the app is depends on the evidence: who published it, how recent it is, and whether it's really on the page.

    PRINCIPLE

    Show how sure you are, honestly

    People rely on a system appropriately when its confidence matches how reliable it really is.2

    Rule 6

    Projections, not promises.

    What this program could do for Nadia

    ~ projection

    ✓

    Opens doors to research and teaching roles

    ✓

    Adds skills she doesn't have yet

    !

    Won't change her pay much at her current level

    Every projection is a range with its sources. And it's honest about what a degree won't do.

    PRINCIPLE

    Say how well it can help

    Make clear what the system can do, and how well.1 A range with a "won't" is more useful than a confident single number.

    Rule 7

    Say when you haven't done the homework.

    RESEARCHED

    Full projection

    Sourced, with a range.

    NOT RESEARCHED YET

    Structural read

    Labeled as such. Never dressed up as analysis.

    No date it can verify? It says "to source." It never fills the gap with a guess.

    Rule 8

    The fit score doesn't use AI at all.

    Fit score

    Plain code

    +

    Open questions

    AI chat

    The fit score has to be right every time, so it's calculated, not generated. AI is added on top for the open-ended questions.

    Rule 9

    Keep her data hers, and name the one exception.

    ✓

    Her profile, résumé and plans stay on her device.

    ✓

    Never sold. Never shared.

    !

    When she asks the AI a question, that question leaves her device. The app says so.

    PRINCIPLE

    Honest beats reassuring

    "Nothing ever leaves your device" would be simpler, and untrue. Tell people exactly what happens to their data.3

    Nadia

    I closed forty tabs. Two programs really move my career, and I can see why.

    BEHIND THE STORY

    Nadia's search was mine.

    I kept applying to program after program and doing all the research by hand. I wanted something that matched my profile to the program that fits my career goals. So I built Pathwise.

    In my words: "I'm done with search engines for browsing education stuff; my app serves my purpose better."

    CHEAT SHEET

    AI that helps people make big decisions

    Decide before you track.

    Match intent, not eligibility. Meaning, not keywords.

    The AI proposes, code confirms, and every fact keeps its source.

    Earn confidence from evidence.

    Ranges, not promises, including what it won't do.

    Label what you haven't researched.

    Name exactly where data goes.

    TRY IT

    Explore Pathwise.

    The live demo walks through a real professional's program search, start to finish.

    EXPLORE PATHWISE →

    You finished case story 03: "Is this degree worth it?" Pathwise is a personal tool by Anuja Nimmagadda, built by directing Claude Code. Nadia and the examples in this story are fictional.

    Sources

    1. Amershi et al. (2019). Guidelines for Human-AI Interaction. CHI 2019 (Microsoft HAX).
    2. Lee & See (2004). Trust in Automation: Designing for Appropriate Reliance. Human Factors 46(1).
    3. Nielsen Norman Group. 10 Usability Heuristics for User Interface Design.