How I’m Using Prompt Engineering to Turn Reports
Into Results
If you’ve ever lost half a day rewriting a status report for
the fourth version of the same slide deck, you’re not alone. It’s one of the
most frustrating parts of the job—and one of the least strategic.
That’s why I started using large language models (LLMs) like
ChatGPT and Copilot. Not just for experimentation, but to systematically take
work off my plate. But here’s the truth:
Generic AI gives you generic results.
Prompt engineering changes the game.
Why Prompt Engineering Matters (More Than You Think)
When most people talk about AI in project management, they
focus on flashy dashboards or automated ticket creation. That’s fine—but it's
not where the real leverage is.
The real impact happens when we combine clean
project data with deliberate prompts—the kind that
generate output we can actually use in executive reviews, steering meetings,
and stakeholder updates.
“You can’t automate what you haven’t standardized.” — Rick
Morris
Prompt engineering isn’t just writing a better question.
It’s a system. And once you build it, the results speak for themselves:
- Weekly
status reports in minutes, not hours
- Risk
logs that auto-rate severity and recommend mitigations
- Steering
decks that don’t require a weekend to assemble
The Three Prompts That Changed My Workflow
Here’s what I actually use week-to-week—no gimmicks, no
fluff:
1. Status Narratives That Make Sense to the C-Suite
I feed in structured data from Jira, Smartsheet, or even a
quick bullet summary. My prompt is tuned to produce something short,
plainspoken, and focused on action.
Key Tip: Ask the LLM to quantify variance and
end with a one-sentence ask.
2. Dynamic Risk Logs
Instead of manually rating every risk, I now have AI assign
RAG status based on impact × likelihood. It also flags missing mitigations or
inconsistent timelines.
Bonus: Add a Markdown or JSON output format so
you can plug it straight into SharePoint or your PMIS.
3. Steering Committee Slide Drafts
I use a slide-friendly prompt that generates headlines,
bullet points, and speaker notes—based on live data from my working files. I
still review and polish, but I’m no longer starting from scratch.
“If you’re doing the same thing every week, it should be
automated.” — Rick Morris
It’s Not Just About Speed. It’s About Focus.
The first time I used prompt engineering to prep a risk log,
I cut my turnaround time by 80%. But the real win? I used those hours to coach
a product owner through a launch delay instead of formatting slides.
Here’s what’s changed in my actual metrics:
- Status
report time: Down from 6 hours/week to ~2
- Risk
updates: From 2 days to under 4 hours
- Exec
clarity: Survey scores on reporting jumped from 3.8 to 4.6 out of
5
“It’s not just about saving time. It’s about reallocating it
to what matters.” — Rick Morris
What You Actually Need to Make This Work
You don’t need an AI team or a budget line. You just need a
repeatable process:
- Know
where your data lives – PMIS, Confluence, Slack threads,
spreadsheets
- Standardize
your outputs – Markdown for risks, JSON for status, PowerPoint
XML for slides
- Build
and iterate your prompts – Make them tight, structured, and
outcome-focused
- Add
guardrails – Validation scripts, human-in-the-loop reviews,
compliance checks
- Track
the impact – Time saved, errors avoided, decisions accelerated
“A good process is one you don’t notice—because it’s
working.” — Rick Morris
Final Thought: The Point Isn’t the AI—It’s the Outcome
LLMs didn’t make me a better project manager. What they did
was give me time back—so I could show up where I was most needed:
in risk meetings, in product launches, in conflict resolution.
Prompt engineering didn’t remove me from the loop. It just
got rid of the noise.
“We don’t manage projects. We make dreams come true.”
— Rick Morris
So if you’re spending more time narrating progress than
driving it, prompt engineering might be your next best move. And if you want to
see my actual prompts or workflows, just ask—I’m happy to share what’s worked
and what didn’t.
Let’s build project management that works for the people doing the work.
