Hilary Gridley’s Playbook for Scaling Yourself with Custom GPTs

By: Hilary Gridley (Head of Core Product, WHOOP) Host: Claire Vo (ChatPRD) Source: ChatPRD - How I AI Type: article (podcast summary)

Summary

Hilary Gridley demonstrates two ways to scale managerial expertise through AI. First, reverse-engineering your own implicit quality standards from before/after examples and encoding them into a custom GPT (“Deck Doctor”) that delivers consistent, criteria-based feedback to your team. Second, using AI as a writing coach through a structured workflow: brain dump → thesis validation → blind spot detection → restructuring — where AI challenges and sharpens your thinking but you always do the final rewrite.

Key Ideas Extracted

Reverse Engineer Judgment Into AI: Collect before/after examples, have AI discover your implicit criteria, then encode into a reusable custom GPT
Be 100x More Specific: Prompting technique that forces AI past vague principles into concrete, actionable standards — broadly applicable beyond the slide evaluation use case
My Job Your Job Role Delineation: Explicit partitioning of human/AI responsibility in a prompt (“MY job is X, YOUR job is Y”) — scopes AI authority, sets output format, preserves human agency
AI as Writing Coach: Structured workflow for using AI to sharpen written communication — thesis validation, blind spot detection, restructuring — where the human always writes the final version
Scale Manager Expertise With AI: Automate the “0-to-60%” repetitive feedback (slide reviews, writing coaching) so managers can focus on deep strategic thinking and personal mentorship

Notes

The “be 100x more specific” prompt is deceptively powerful — it’s a general-purpose technique for forcing AI past platitudes into actionable specifics. Works for any domain where you need concrete criteria.

The reverse-engineering approach is interesting because it starts open-ended intentionally — she doesn’t want to bias the AI with her own framing first. Let AI discover patterns, then human refines. This is the opposite of most prompting advice (“be specific upfront”).

The writing coach workflow embodies a key principle: use AI to challenge your thinking, not to replace it. The AI should help you “make it right” not “validate that it’s right.”

Raw Content

Hilary Gridley, Head of Core Product at WHOOP, shares how she builds custom GPTs that clone her expert judgment to evaluate slide decks and uses AI as a writing coach to help her team’s ideas go viral.

Key Philosophy

The goal isn’t to replace managers but to give them superpowers — automating repetitive feedback so you can spend time on deep strategic thinking and personal mentorship.

“I kind of want the AI to start by interpreting this in ways that I might not even be able to predict. And then I’m gonna get an intune it and that’s when I get super, super specific.”

“Try to get the AI to help you make it right as opposed to assuming that it’s right and getting the AI to validate that.”

Workflow 1: Building a Custom GPT to Think and Evaluate Like You

The problem: as a manager, explaining what “good” looks like is hard. You know it when you see it, but turning intuition into clear, consistent feedback is a challenge.

Step 1: Collect ‘Good’ and ‘Bad’ Examples

Two-column document of ‘before’ and ‘after’ slides
Save as PDF — raw training data for AI to understand your taste

Step 2: Reverse-Engineer Your Criteria with AI

Upload PDF to ChatGPT with intentionally simple, open-ended prompt (to avoid biasing the AI)
AI analyzes before/after pairs and generates first-draft list of principles

Step 3: Refine and Get Hyper-Specific

Key prompt: “Be 100 times more specific.”
Forces AI beyond vague principles into concrete, actionable standards
Blend AI pattern-matching with your expert judgment

Step 4: Build the ‘Deck Doctor’ GPT

Have the AI write GPT instructions: “MY job is to create a GPT that can evaluate slide decks… YOUR job is to write the prompt for it.”
AI generates detailed instructions including persona details
Copy into GPT’s ‘Instructions’ field

Step 5: Test and Deploy with Your Team

Team members upload decks and get instant feedback based on your standards
1-5 rating on each criterion, concrete feedback, improvement suggestions
Takes “0-to-60%” feedback off your plate

Workflow 2: Using AI as a Writing Coach to Make Your Ideas Go Viral

Thesis: the best way to get invited to important strategic meetings is to be pulled in by writing up a strong point of view that gets shared around.

Step 1: Brain Dump and Sanity Check

First draft, no filter
“Can you succinctly express my thesis back to me and my main supporting points?”
If AI can accurately summarize → on track. If not → clarify thinking first.

Step 2: Make It More Compelling

“How can I make this more compelling?”
AI as sparring partner — not rewriting, but expanding your thinking

Step 3: Identify Blind Spots

“What blind spots might I have as I’m talking about this?”
Flag: over-indexing on one user type, underplaying structural forces, false equivalences

Step 4: Restructure for Clarity and Final Polish

Ask AI to suggest new structure
Key principle: the process always ends with YOU
Take all AI feedback → do the final rewrite yourself

Takeaways for Managers

Build expertise into custom GPTs → coaching tools that scale endlessly
Use AI as writing partner → sharpen thinking, bigger impact
Automate repetitive feedback → free yourself for deep strategic work
The “be 100x more specific” technique is broadly applicable