Wednesday, September 16, 2026
AI for HR: Practical Tools, Better Prompts, Real Use Cases
Your companion site for the day and after.
Prompt library
The prompts live on their own page
The library is long, so it has its own page with copy buttons on every card, grouped by Communications, Benefits, Compensation, Talent, Editing, and Building your own tool.
Hands on
Exercises
Three exercises that match the workshop flow. Pick the one that fits what you came to do.
- 01
Improve one prompt with FRAME
Take one real thing you have to write this week and rebuild the prompt using the five FRAME steps.
- 02
Build sprint
Start building the tool you selected at registration in Lovable. First milestone: get to a working input box before anything fancy. Ambassadors are in the room to help.
- 03
Watch and plan (for observers)
Using the idea-to-shipped path, sketch the inputs, outputs, and edge cases for a tool you wish existed. Leave with a plan even if you do not build today.
Reference
The builds
Start from the idea-to-shipped path, then pick something from the build menu.
The idea-to-shipped path (seven steps)
- 01
Name the dependency. What, or who, are you waiting on? Write it in one sentence. If you cannot name it, you are not ready to build.
- 02
Front-load the reasoning. In a regular AI chat, think the whole thing through before you touch a builder: what are the inputs? What is the output? What are the edge cases? What data does it need, and where does that data live? Cleaner thinking here means far fewer messy iterations later.
- 03
Let one AI write the build brief. Have the chat model turn your reasoning into a clear brief you can hand to the builder.
- 04
Build in small moves. Paste the brief into the builder, then work in small, testable steps. Small steps are easier to fix than big ones.
- 05
Test with real, messy data. Not the clean example. The ugly real thing. That is where it breaks, which is exactly what you want to find before anyone else does.
- 06
Audit with HUMAN. Run the output through the Human Stewardship Model. Facts, math, tone, bias, policy fit.
- 07
Ship to one real person. Hand it to a single colleague who has the actual problem. Watch them use it. Then widen. Building the tool is the easy part; earning trust in it is the real work.
Build menu
Four worked examples across compensation and total rewards, benefits and employee experience, and talent and HR operations.
Compensation and Total Rewards
Job Description / Job Architecture Matcher
Removes the dependency on a comp expert to level a role.
You paste a job description; the tool maps it to your leveling framework and flags where it does not fit cleanly. The payoff is not just speed; it surfaces roles that were quietly leveled wrong. The expertise gets encoded once so a recruiter or HRBP can use it without you in the room.
Compensation and Total Rewards
Comp Offer Tool
Removes the dependency on manual, error-prone offer math.
The best version of this tool has no AI running inside it: AI helped design and build it, but at runtime it is just clean, deterministic logic. When the answer has to be exactly right every time, you do not want a model improvising.
Benefits and Employee Experience
Policy Explainer Chatbot
Removes the dependency on HR to answer the same policy question fifty times.
Grounded in your actual handbook, not the open internet, it answers employees in plain language and points them to the source.
Talent and HR Operations
Interview Question Generator
Removes the dependency on each hiring manager to write good, role-specific questions from scratch.
You give it the role and what you are screening for; it generates structured, consistent questions grounded in that role. It drafts the questions, but a human still evaluates the answers.
Governance
Human stewardship, step by step
Two graphics and two checklists you can run through before you paste anything in, and before you ship anything out.
Before you paste anything in
Before you ship anything out
This site is educational and is not legal, tax, or compliance advice.
Setup
Tools and setup
Use the accounts you already have. Just observing is completely fine; you do not need to sign up for anything.
- Open
ChatGPT
General assistant: drafting, rewriting, analysis.
- Open
Claude
Strong on long documents and careful reasoning.
- Open
Gemini
Google's assistant, handy alongside Workspace.
- Open
Copilot
Microsoft's assistant inside Office and Teams.
- Open
Lovable
Build a working tool by describing it. Free credits provided for the workshop; the paid upgrade is optional.
- Open
Gemini Notebook
Ask questions across your own documents. Free with a Google account.
Leave with a plan
My action plan
Fill this in during the last stretch of the workshop, then print or save it. Nothing is stored and no account is needed.
Nothing you type is stored or sent anywhere; it stays in your browser.