1Understand evals in plain English
Learn what an eval is: a test with answers you already know.
live workshop · rsvp open
How to build your first AI Eval
Nate Grahek AI Educator @ The Rundown University
Learn how to measure AI quality, compare models, and trust workflows before you scale them.
- When
- Thursday, August 27, 2026 · 2:00 PM EDT Thu, Aug 27 · 2:00 PM EDT
- Length
- 1 hour 30 minutes
About this live course
Too many AI rollouts still can't answer the money question. It's rarely the model. It's that nobody defined what good looks like, so nobody can say whether the thing works, what it costs per result, or whether last month's tweak made it better or worse. The organizations seeing real ROI from AI are the ones measuring it. That's not a coincidence.
This workshop is a guided tour of your first evaluation. We'll take a spreadsheet of messy inbound leads, have AI score and prioritize them in Claude Code, then grade the results against answers we already know are right: which deals closed and which went nowhere. You'll leave with an eval-first way of thinking, a simple way to run your first eval, and the question to ask before trusting any AI workflow with real volume.
This is an intro session. If you can read a spreadsheet, you're qualified.
Builders leave with a repeatable pattern to try on their own data. Leaders leave with the one question to ask before green-lighting an AI project. Consultants leave with a way to show clients a number instead of a feeling. The same method works in OpenAI's Codex too, and we'll show you where.
In this session, you'll learn
5 outcomes2Build a first eval in Claude Code
Watch a live build that scores and prioritizes messy inbound leads, with no code-writing required.
3Measure the work that creates ROI
See why AI initiatives that show ROI are the ones that define and measure success.
4Compare accuracy and cost
Weigh models against each other on result quality and cost for the same job.
5Build toward self-improving AI
Use evals as the foundation for AI systems that can improve their own work.
What to expect
Live on YouTube
Come back to this page when the session starts and join with one click.
Hands-on workflows
Real setups you can copy, not slideware.
Open Q&A
Bring the AI workflows you are trying to make more dependable.
Replay for 48 hours
The recording lives on this page for 48 hours after the session, with an AI summary PDF.
A preview of what we'll cover
- A plain-English understanding of AI evaluations.
- A repeatable eval pattern you can try with your own data.
- A practical way to compare model accuracy and cost.
- The question to ask before trusting an AI workflow with real volume.
- A clear path for applying the same method in Claude Code or OpenAI Codex.
Tools used in this session
Claude CodeOpenAI CodexSpreadsheets
Your instructor
Nate Grahek AI Educator @ The Rundown UniversityNate is a SaaS founder and Fractional CMO who helps product-driven businesses build marketing systems that actually work — without the fluff. He's spent years helping founders and operators cut through marketing complexity and put the right things on autopilot.
At Rundown University, Nate brings that same hands-on, no-jargon approach to AI education. His workshops and courses focus on practical automation and AI workflows you can deploy the same day — no engineering background required. If you've ever wanted to use AI to get your time back, Nate shows you exactly how.
