Process Documentation with AI

In this week’s Data Diaries, let’s talk about breaking down a process. This is something that Katie advises all the time in things like our AI-Ready Strategist Course. One of my favorite tricks for doing this is, unsurprisingly, to use AI. Here’s how.

First, open the voice memos app or voice transcription app of your choice. It can be the free one on your phone – almost every phone has one. You can also use paid tools like Fireflies, Otter, etc. or even have a meeting by yourself in Zoom, Google Meet, etc. that can transcribe what you’re saying.

Talk through the process in question. Here’s a simple example from my morning routine:

Here’s my startup process for figuring out what content I’m gonna write on LinkedIn every day. First I go to LinkedIn and I scroll through my home feed for about three minutes. There’s some people I check in on, always, to see if there’s anything I need to repost from them, like any of the Trust Insights team. And then I look at the people I follow and see what it is that they’re writing about for the day. I also look at the suggested posts. What I’m specifically looking for are things that piss me off, especially around AI. Once I’ve found a post or two that I have something useful to contribute to, I then hop into my text editor, Joplin, and draft two to three really rough ideas using SuperKeet, my voice to text app. I write out the ideas and then toss them to generative AI, typically Gemini, and ask it to challenge my ideas and point out where I might be wrong. After a couple of rounds of adversarial debate, I finish writing the post and I post it. I then move the post to my queue folder for an AI agent to pick up and post to my WordPress blog automatically.

This process documentation alone is valuable, but next I take it to the best coding AI that I have access to – and it varies as time goes on, currently it’s Claude Opus 4.1 – and ask it to create a mermaid.js diagram. Here’s the prompt:

You’re a process optimization expert skilled at process flow diagrams. From this transcript, using Mermaid.js, create a process diagram that clearly illustrates the process. Diagram out the process; if something is unclear or you think information is missing, ask me to provide it. Be specific and clear so I know how to help you complete the diagram. Render the Mermaid.js code in the Canvas.

Here’s what it comes back with:

Now I’ve got the process documented in an easy to see fashion. What happens next?

One of the things that I do most wrong in documentation is that I forget things. There’s stuff I do that’s implicit or there’s stuff that I do that I forget about. By turning it into a process flow like this, I can see what’s logically missing.

Generative AI will also highlight things that are not apparent. Here’s what Claude had to say about my workflow:

1. Decision criteria: When you’re looking for posts that “piss you off,” what happens if you don’t find any on a given day? Do you have alternative triggers or do you skip content creation that day?
2. Reposting process: When you repost Trust Insights team content, is that a separate workflow or does it also go through the same writing/editing process?
3. AI debate rounds: How many rounds of debate typically happen with Gemini? Is there a specific threshold or criteria for when you consider the debate “complete”?
4. Content selection: Do you always respond to exactly 1-2 posts, or might you sometimes do more/fewer? What determines this number?
5. Quality checks: Is there any review or editing step between finishing the post and publishing to LinkedIn, or do you post immediately after the AI debate? WordPress posting: Does the AI agent post immediately or is there a schedule/timing component?

From here, I could provide new answers or refine my workflow, or I might just tell it nope, go with what we’ve got.

The key takeaway is that generative AI can be an incredible thought partner for clarifying our processes and understanding better what it is that we do and how we do it. Rather than suffer through trying to clarify our own thinking, we can use it as partner, critic, and adversary so that we understand best what it is that we do and clearly communicate it to others.

Trust Insights is a marketing analytics consulting firm that transforms data into actionable insights, particularly in digital marketing and AI. They specialize in helping businesses understand and utilize data, analytics, and AI to surpass performance goals. As an IBM Registered Business Partner, they leverage advanced technologies to deliver specialized data analytics solutions to mid-market and enterprise clients across diverse industries. Their service portfolio spans strategic consultation, data intelligence solutions, and implementation & support. Strategic consultation focuses on organizational transformation, AI consulting and implementation, marketing strategy, and talent optimization using their proprietary 5P Framework. Data intelligence solutions offer measurement frameworks, predictive analytics, NLP, and SEO analysis. Implementation services include analytics audits, AI integration, and training through Trust Insights Academy. Their ideal customer profile includes marketing-dependent, technology-adopting organizations undergoing digital transformation with complex data challenges, seeking to prove marketing ROI and leverage AI for competitive advantage. Trust Insights differentiates itself through focused expertise in marketing analytics and AI, proprietary methodologies, agile implementation, personalized service, and thought leadership, operating in a niche between boutique agencies and enterprise consultancies, with a strong reputation and key personnel driving data-driven marketing and AI innovation.