The Operator's AI Toolkit
Field Manual: Practical tools and guidance to identify high-leverage AI use cases
Your team is drowning in operational chaos. People are spending hours crafting the same types of emails over and over. Your best performers are buried in administrative tasks instead of strategic work. Critical decisions sit in bottlenecks waiting for approval.
This is exactly when AI can help most- not by adding more complexity, but by removing the friction that's holding your team back.
In my last post, we established that your AI usage is a workflow decision, not just plugging in the latest tech. This Field Manual delivers the what to do after the how to think.
Here’s the process we’ll walk through:
Table Stakes: Are You Ready to Start?
Before experimenting with AI, make sure you have the basics:
✅ Your team collaborates digitally (email, cloud storage, video calls)
✅ Most of your files are digital, not paper-based
✅ You have basic security practices in place (i.e. password management and an understanding of what confidential data is)
✅ Leadership and team willingness to potentially briefly slow down in order to speed up
If you're still faxing documents or keeping everything in filing cabinets, focus on digital transformation first.
AI Experiments: Understanding What AI Can Do
As Stephen King said, "The scariest moment is always just before you start." Staring at that blank chat screen can be daunting, so here are some experiments to demo common use cases.
Team Strengths Matrix: Where AI Helps Most
Now that you've seen what AI can do through the experiments, let's figure out where it will have the biggest impact on your team.
Research from MIT, Harvard, and Stanford shows AI elevates mid-performers more than high-performers:
Now apply this to your team: Create a strengths matrix to identify where AI will have the biggest impact.
Building Your Team's Strengths Matrix
Step 1: List your core business functions at the right level of granularity:
Step 2: Plot each function on this matrix. Business importance can be literal (customer on-boarding) or be a domino that limits top performers (Closing sales is great, but the team spends too much time writing template proposals).
Step 3: Focus AI adoption on the "Strengthen These" quadrant first - high business importance where your team has room to grow. Your strongest capabilities in the "Not a lot of AI value" quadrant won't see huge AI gains. Your struggling capabilities in "Strengthen These" will.
Friction Mapping Exercise
You've tested AI's capabilities and mapped where your team could benefit most. Now let's identify the specific pain points where AI can make a difference.
Instructions:
Gather 2-3 team leaders for 30 minutes
Use these prompts to identify where time, attention, or quality suffers:
🔍 Key Questions:
Where are we repeating ourselves every week?
Where do decisions get bottlenecked because everything flows through leadership?
Where do things fall through the cracks due to fatigue or bandwidth?
What would we do more of if it were 30% faster?
What's something we stopped doing because it became too time-consuming?
📋 Output: A list of 5-7 friction points where AI might help
Use Case Brainstorm: From Friction to Function
Now you have the full picture: what AI can do, where your team needs help most, and what's causing friction. Time to connect the dots.
Turn your friction points into testable AI use cases:
Step 1: Take your friction mapping results and consider them in the context of your strengths matrix
Step 2: For each friction point, ask things like:
Could AI create a first draft if a human reviews the final version?
Would success here build momentum or unlock other improvements?
Step 3: Score each potential use case:
Step 4: Select one or two use cases to pilot. Start with high impact, easy implementation, and low risk. For each one, think about what the outcome should be—greater efficiency, time reduction, etc. Know what success looks like before starting the pilot.
Security & Privacy: The Non-Negotiables
Before you start implementing any of these use cases, let's make sure you're doing it safely.
❌ What NOT to put in AI tools:
Customer personal information
Financial data or passwords
Confidential strategic plans
Legal documents without review
✅ Tool selection matters:
Consumer tools (ChatGPT, Claude): Convenient but data handling varies
Enterprise tools (Copilot, enterprise versions): Better privacy controls but you may not have licensing
Always ask: Does this tool train on my data?
📋 Meeting transcription privacy considerations: Meeting transcription is the most common AI use case people encounter, but there are important privacy concerns to watch out for:
Tools that currently don't train on your data: Fireflies, Microsoft Copilot (enterprise)
Tools that currently use data for training: Otter (though de-identified and with consent)
Essential steps: Always get explicit consent before recording. When in doubt, anonymize
Key principle: Choose tools based on your privacy requirements
Resources for Deeper Learning
If you want to invest more time in learning about deeper AI capabilities:
Understanding AI Capabilities:
Co-Intelligence by Ethan Mollick - Practical AI insights
AI Deep Dive Course on Udemy
Prompting Guide and Cheat Sheet from Superhuman.ai
Privacy & Security:
Tool-specific privacy policies (read them!)
Your IT team's AI guidelines (if available)
Start Where You Are
You now have the tools to identify where AI can reduce friction in your organization without adding chaos. The goal isn't to implement AI everywhere- it's to find the specific places where it can free your team to do their best work.
Start with one experiment this week. If it works, build from there. If it doesn't, try another. The path to driving outcomes with new technology when you don’t have an endless innovation budget is to test rapidly and learn.
- Bryan