Session 2 · Everyday AI Use

The Vibe Lab

An interactive vibe-coding session

Thursday, September 24, 2026  ·  KRESA Career Connect Campus  ·  Hosted by KalFlow

Before You Arrive

Bring a laptop and leave with a repeatable process. In this hands-on session you’ll take a messy year of IT service tickets and use AI to review it, catch the errors hiding in it, and turn the whole thing into a recipe you can run again next month on your own data. No coding experience required — and no prior session required either.

Agenda

  • 01
    Introductions & Polling Questions

    Meet the room — who’s back from June, who’s new — and find out where everyone is on their AI journey.

    ~15 min
    • Name, company, role, and current AI experience level
    • Which tools you’re using — Claude, ChatGPT, Grok, or something else
    • What you’ve put AI to work on since we last met, and what’s still tripping you up
  • 02
    Lab Overview

    A quick tour of this session’s dataset, the two paths you can choose from, and getting everyone set up before we start.

    ~10 min
    • Pick your track: Operations Support or Reporting & Insight
    • Open the data files emailed to you ahead of the session
    • Set up a folder so today’s work is reusable after you leave
    • No AI account yet? We’ll get you on a free plan in about two minutes
  • 03
    Getting Repeatable Results

    The habits that separate a lucky answer from a result you can count on every time — treat AI like a brilliant new teammate with zero context, and do the onboarding.

    ~20 min
    • Five habits: give it a role and a goal, be specific, ground it in your data, review before you act, capture what worked
    • The four-step loop that works on almost any task
    • Where this leads: a one-off chat → a reusable recipe → a saved skill
  • 04
    Interactive Lab

    Your turn. A new dataset, the same four-step loop — twelve months and 161 IT service tickets, some of them miscoded, mis-billed, or carrying impossible dates. On purpose.

    ~40 min
    • Prompt the AI to find what’s wrong, then either clean it up or report on it
    • Discover the errors you’d never catch by scrolling — the ones that only show up when you compare one column against another
    • Capture the process as a document you can reuse in a fresh chat anytime
    • Group debrief: what your AI caught, what it missed, and where you’d run this loop at work
The Loop You’ll Practice
  • STEP 1
    Review
    Look, don’t touch. What do you see?
  • STEP 2
    Validate
    Check the rules; flag what’s wrong.
  • STEP 3
    Act
    Clean it, summarize it, or chart it.
  • STEP 4
    Capture
    Write down the process to reuse.

Choose Your Track

Operations Support

Produce a cleaned copy of the ticket log with flags and suggested fixes — marked for human approval, never silently overwritten.

Reporting & Insight

Build charts of billed revenue by service type and monthly ticket volume, then see exactly how the uncaught errors would have distorted them.

What to Bring

  • A laptop — this is hands-on, not a demo you watch
  • An AI account. Our demo runs in Claude, but every prompt works in ChatGPT or Grok too. Don’t have one? Arrive a few minutes early and we’ll set you up free, no credit card.
  • A real work problem in the back of your mind — you’ll want somewhere to point this loop on Monday

New to the Vibe Lab? You’re in the right place. Session 2 uses a fresh dataset and stands on its own — returning attendees are repeating the loop on new data, which is exactly the point. We keep the group small so the lab stays hands-on.