Maya, a Senior People Ops Manager, spent every Sunday night the same way for three years: open her laptop, try to remember what she accomplished the previous week, write her weekly update for her boss, and plan her week ahead.
Then, on Monday, she'd open her laptop again, full of optimism that she was going to accomplish all her goals... and immediately fall down a Slack rabbit hole.
2-hours in, she didn't even know what she was supposed to be working on.
Meanwhile, her CTO had been pushing the team toward AI tools for months. Maya had nodded along, installed a few things, used them occasionally. She assumed she was keeping up. She wasn't.
The shift happened when she stopped trying to "use AI more" and started asking a specific question based on her pain point: How can I make it easy to know what to work on today?
She spent an hour one day connecting Claude to her task manager, her calendar, Slack, and her AI meeting notes tool. She set a few rules, like always protect a lunch window between 12 and 2pm, surface anything flagged as urgent, and assign remaining tasks to calendar blocks by priority. Then she let it run.
That was five weeks ago. Here's what Maya's week looks like now.
Every morningโ
As Maya grabs her coffee, she opens Claude so it can run its daily scheduled routine. It reads her quarterly priorities, her calendar, pulls her tasks, scans overnight messages, and reviews meeting notes from the previous day. By the time she sits down to work, she has a prioritized schedule with lunch protected, urgent items surfaced, email and Slack responses drafted, and one most important "anchor" task mapped to her longest open time block.
She doesn't decide what to work on first. That decision was already made for her, with her guidance, in advance.
Every Mondayโ
A Claude routine drafts two things: her weekly update for her manager and a batch of celebration posts for team members. The routine pulls from Slack messages, completed tasks, and meeting notes to surface wins she'd otherwise forget to recognize. She reviews, edits, and sends. Total time: under 10 minutes.
Every Wednesdayโ
A routine reviews her notes from the previous week's 1:1s with her direct reports as well as HRBP sessions with people managers and delivers a research-backed coaching brief: where she showed up well, where she missed something, what patterns are emerging across her team. Like a weekly debrief with a coach who was in every meeting.
Every Fridayโ
A routine reads what happened that week โ tasks completed, calendar events that occurred โ and updates her goal progress automatically. If she committed to 12 stay interviews this quarter and 3 were on her calendar this week, her tracker increments by 3. No manual logging. No end-of-quarter scramble trying to remember what she did in April.
By week three, Maya stopped asking herself if she was "an AI person." She had her answer. And she was so excited to keep exploring.
Are you working differently today than you were a year ago?
Maya's story is fictional. The routines are not. They're real, and various versions of them are being run by HR professionals around the world.
If you ran your current workflows for another year without changing anything, would you be ahead or behind where you are today?
In most workplaces, you'll be behind. Not because you're bad at your job, but because the floor is moving, and staying still is its own kind of falling.
That in and of itself can be anxiety inducing. And, as human beings, we rarely do our best work from a place of anxiety.
So, try to shift your perspective by finding one specific thing related to AI that makes you genuinely excited or curious, and go build or try that thing.
For me, it was building my own website... with Claude Code... with little to no prior web development knowledge. You can check it out here.
I wrote about the step-by-step process on LinkedIn. The response surprised me. People shared they felt like this was the future. And, they were already working towards that future with similar builds themselves.
Here are the types of things HR professionals are experimenting with:
Before you can build it, you need to understand it
Most HR professionals I talk to are curious about AI but don't know where to start, partly because the vocabulary is specialized.
Here's the short version for Claude (which is where I'd recommend you start, but Lovable and Replit are next if you're already vibe coding or building apps).
What is Claude Chat/Cowork/Code?โ
Claude Chat/Cowork/Code is Anthropic's AI environment where Claude is your active collaborator. It can read your files, connect to your tools, and execute tasks, all from plain-English instructions. You don't need to know how to code. As you move from Chat to Code, Claude can do increasingly sophisticated and independent work. You can also influence this with which model you select (e.g., Sonnet for everyday tasks vs Opus for deep work).
What's a skill?โ
A skill is an on-demand capability you invoke when you need it. "Draft my email replies." "Write my manager update." You run it, you get an output, you're in control of when it fires.
What's a routine?โ
A routine is a scheduled task that runs automatically on a cadence you set. Your morning planning routine fires every morning whether you think about it or not. Your Friday goal-update runs at 5pm every Friday. You build it once.
A routine will only run while your computer is awake unless you're using something like a Mac Mini (that's why you're seeing so many more photos of tech bros showing stacks of Mac Minis on their desks. It's the new flex.)
What's an MCP?โ
MCP stands for Model Context Protocol. It's what allows Claude to actually connect to your tools to read from your task manager, write to your calendar, and search your messages. Without MCPs, Claude is a smart assistant that can only see what you paste in front of it. With MCPs, it can see and act across your entire workflow.
What are connectors?โ
Connectors are pre-built MCP integrations. Instead of writing code to connect Claude to Gmail, you install the Gmail connector. They're the bridges between Claude and the tools you already use.
The combination of skills + routines + connectors/MCPs is what separates "I use AI to write emails sometimes" from "AI is embedded in how I work every single day."
The Four-T Playbook
Every edition, I share a proven tip, trick, tactic, or template. This time, it's a:
๐ Tip: Your 30-day AI workflow transformation.
Treat it like a 5K training plan. (Hilary Gridley actually offers a Couch to 5k for AI plan.) You're not running a marathon in month one. You're proving to yourself that you can run consistently and building the habit that makes everything else possible.
This week: Sign-up for Ruben Hassid's How to AI newsletter. Install Claude. Create your Claude.MD context file. Set up one connector to a tool you already use.
Week 2: Pick the one workflow that, if it ran automatically, would save you the most time or mental load. (Or, if AI is overwhelming to you, just pick something that you'll get excited about). โ
โ
Week 3: Build and test your first skill or routine. Run it manually a few times. See what breaks. Iterate.
Weeks 4: Let it run. Use it. Notice what changes, not just in the time you save, but in what you're able to focus on instead.
At the end of 30 days, ask yourself the question again: Am I working differently today than I was a month ago?
If you can say yes, you've cracked it. Everything else is just adding more rooms to a house you now know how to build.
Quick note: AI adoption policies differ across companies. Before purchasing or connecting Claude to tools that handle employee data, check your company's acceptable use and data privacy policies and loop in your legal team as needed.
Final thoughts
The world is moving fast. The advantage is shifting toward people who can adapt quickly. That's you... if you decide it is.
I'd love to hear from you: How are you going to start working differently with AI this month? Hit reply.
Until next time,
Melissa
P.S. I've got one spot open for bi-weekly to monthly one-on-one mentorship. Learn more and apply here.