LENS Learning Engineering for Next-Generation Systems
Prototypes

Problems → Prototypes · AILE

← Back to this problem brief

An AI Academic Navigator for Two-Year Colleges

Students in open-access and two-year institutions often navigate fragmented academic, financial, technology, and support systems while balancing work and family responsibilities. Generic AI can provide 24/7 help but may confidently invent institutional rules or conceal where human support is necessary.

Try it

This exercise runs locally using simple rules. Your entries stay on this page and clear when you reload. No AI service is called.

Local practice controls

Fictional college navigation practice

These authored cards are approved only for this exercise. They describe no real institution. This tool selects cards and prepares questions; it cannot answer institutional policy questions or contact an office.

F1 · Fictional catalog

The published course schedule lists meeting times and delivery format. An advisor helps learners check program sequencing.

F2 · Fictional enrollment guide

Before changing enrollment, ask advising about progress and the aid office about any funding implications. These offices provide the binding answer for the learner’s circumstances.

F3 · Fictional access services guide

Access services explains its request process privately. Do not put personal records into this practice tool.

Use invented or non-identifying details. Nothing is saved automatically. Downloads contain your text; you decide whether to share them.

Critical review of this working prototype

Strongest assumption: Visible fictional source cards improve safe self-navigation compared with a static office directory.

Likely failure: A schedule selection can hide an enrollment or funding question; readers may mistake practice cards for actual college rules.

Test that could change the design: Give learners ambiguous schedule/drop questions and an office-directory control. Reject the design if learners act without verification or cannot identify the responsible office.

Non-AI comparison: An accessible printed office directory with example questions and phone/in-person routes.

Evidence to collect: Observe correct referral, teach-back of next action, and later successful human follow-through; clicks are not learning. No telemetry is collected.

Human control and access: Keyword routing can miss paraphrases and other languages. All outputs retain source scope and human verification; no institutional policy, eligibility, or clinical decision is made. Downloads are user-controlled and may include private text.

Changes made during review

  • Added sensitive-question override independent of topic selection.
  • Kept all fictional approved sources visible before use.
  • Added unsupported and wellbeing routes, verification checklist, and understanding prompt.
  • Disabled all form controls in server markup until local handlers are installed, preventing no-JavaScript GET submission of entries.
  • Restored the full eight-step path from shared metadata without changing shared files.
  • Smallest next experiment: observe a human-reviewed teach-back using an unseen institution directory.

This is a design and implementation review, not an empirical validation of learning outcomes.

Optional AI critique

Take the brief into your own AI environment

Nothing is sent until you choose.

Copy a self-contained review prompt, then open the AI workspace you already use. The prompt asks for a falsification test, a non-AI alternative, evidence that distinguishes activity from learning, and human-system risks.

Open ChatGPT ↗ Open Claude ↗ Open Gemini ↗

Preview the prompt 
Act as a critical learning-engineering reviewer. Review the prototype brief below as a thin-slice experiment, not as a finished product.

Return:
1. The strongest design assumption.
2. The most important plausible failure mode.
3. The highest-value falsification test for the next cycle.
4. One credible non-AI alternative that could address the same capability gap.
5. The next smallest prototype worth building.
6. What should be measured to distinguish engagement from learning, transfer, and real performance.
7. Human-agency, equity, accessibility, privacy, or governance concerns that should change the design.

Be concrete. Separate evidence-backed claims from hypotheses. Do not reward novelty for its own sake, and do not assume more AI is better.

COLLECTION: AILE
TITLE: An AI Academic Navigator for Two-Year Colleges

CAPABILITY GAP:
Students in open-access and two-year institutions often navigate fragmented academic, financial, technology, and support systems while balancing work and family responsibilities. Generic AI can provide 24/7 help but may confidently invent institutional rules or conceal where human support is necessary.

FIRST PROTOTYPE:
Create a retrieval-grounded academic navigator using only approved institutional sources. It explains options, checks understanding, and escalates financial aid, disability, mental health, or policy-sensitive questions to people rather than making decisions.

REPRESENTATIVE USE CASE:
A working adult asks how dropping one course might affect progress. The navigator identifies the relevant institutional resources, explains which variables matter, and routes the student to the appropriate office for a binding answer.

LEARNING / HUMAN-SYSTEM FRAME:
The learning goal is institutional self-navigation: knowing what to ask, how to interpret information, and when to seek human help. Human-centered AI design requires calibrated confidence and clear boundaries.

QUESTIONS ALREADY IDENTIFIED:
1. Which questions are safe to automate?
2. Does the navigator improve successful follow-through, not merely chat satisfaction?
3. Are nontraditional learners using and trusting the system equitably?

FIRST-CYCLE FRAMING:
Understand: Validate that the real gap is the ability to navigate institutional systems and make informed next-step decisions, not simply low engagement, low tool use, or a workflow inconvenience.
Map: Map the system: student, advisors, institutional services, approved knowledge base, AI navigator, and escalation workflow. Identify where the capability currently succeeds, breaks down, or is masked by other constraints.
Instrument: Predefine evidence: correct action, successful referral, resolution time, hallucination rate, trust calibration, and access by learner group. A key disconfirming signal is: usage is high but students still miss deadlines, choose wrong actions, or avoid human support when needed.
Open the underlying design brief and eight-step path

A first prototype

Create a retrieval-grounded academic navigator using only approved institutional sources. It explains options, checks understanding, and escalates financial aid, disability, mental health, or policy-sensitive questions to people rather than making decisions.

Representative use case

A working adult asks how dropping one course might affect progress. The navigator identifies the relevant institutional resources, explains which variables matter, and routes the student to the appropriate office for a binding answer.

Questions worth carrying forward

  • Which questions are safe to automate?
  • Does the navigator improve successful follow-through, not merely chat satisfaction?
  • Are nontraditional learners using and trusting the system equitably?

Evidence anchors

These sources motivate the design; they do not validate this prototype.

LENS iteration cycle

One possible first pass

Refine → Understand
  1. 01
    Understand

    Validate that the real gap is the ability to navigate institutional systems and make informed next-step decisions, not simply low engagement, low tool use, or a workflow inconvenience.

  2. 02
    Map

  3. 03
    Design

    Compare an AI intervention with simpler non-AI options. Specify what the human decides, what the AI may suggest, and which tradeoffs are acceptable.

  4. 04
    Build

    Create the smallest usable prototype around one representative task, with expert-curated content, visible uncertainty, and an easy human override.

  5. 05
    Instrument

    Predefine evidence: correct action, successful referral, resolution time, hallucination rate, trust calibration, and access by learner group. A key disconfirming signal is: usage is high but students still miss deadlines, choose wrong actions, or avoid human support when needed.

  6. 06
    Deploy

    Pilot with a small, representative group in a near-real setting; preserve a baseline or comparison condition and document implementation conditions.

  7. 07
    Evaluate

    Look for capability growth and transfer, not just satisfaction or activity. Inspect subgroup patterns, human workload, errors, and unintended adaptations.

  8. 08
    Refine

    Let the evidence change the problem model. Default back to UNDERSTAND if the assumed gap was wrong; otherwise revisit the earliest step invalidated by the evidence.